From libssh2-devel-bounces@cool.haxx.se Mon May 2 00:22:02 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p41MLcC4020744; Mon, 2 May 2011 00:21:58 +0200 Received: from mail-px0-f182.google.com (mail-px0-f182.google.com [209.85.212.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p41MLZsT020603 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Mon, 2 May 2011 00:21:37 +0200 Received: by pxi20 with SMTP id 20so1608922pxi.13 for ; Sun, 01 May 2011 15:21:37 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:date:message-id:subject:from:to :content-type; bh=gYESgrrVtkVRA8aqYe4+kstnP1Qx4h31aXYep0QuXTc=; b=YzgYJ7secbc3LcDnF+pYPGqQUfEu0/A7E9Y7TSzvX3uM9BRC2v2SiLxLDePoph/UUm tyJRlYbXDNHjUf1FqaT/5DWwZH0Bn3qtqgcEm0FdDH1YcX7pK/Tl9aWOJXiJGPXge0kS 5mxjMFGh8JhtQvCNdaKEZDcGCMZhEP8Gnxo0U= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=tTHvc7Am5YrxHN8iBly/XmWEv0elxCHyUFIgt8DAPLz12tuxZb888xkAcZ9WtwLYxT H9Zz7S6baL2PoLQjUt48CzsmE3fImZCRymQy6DYwagLKlKGenGDobQ1I7Xv9BLWkhUl2 XYqZzMYwB2d33PoEi9tiEF8bbMEPUnolqZajA= MIME-Version: 1.0 Received: by 10.68.8.225 with SMTP id u1mr5408439pba.409.1304288497300; Sun, 01 May 2011 15:21:37 -0700 (PDT) Received: by 10.68.40.198 with HTTP; Sun, 1 May 2011 15:21:37 -0700 (PDT) Date: Sun, 1 May 2011 15:21:37 -0700 Message-ID: Subject: Attempting to wrap sftp in Qt From: Mark Roden To: libssh2-devel@cool.haxx.se X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 00:22:02 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 00:21:37 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Hi all, I'm getting an error with trying to copy a file using sftp with libssh2 from a windows 7 machine to a linux machine using libssh2. The error number is 2, and from this thread here: https://bugs.launchpad.net/paramiko/+bug/492238, there's a suggestion that it might be a permissions error. But it's still a problem, and I'm hoping for some help here. My test code looks like this: SSHFunctions theFunctions; QString hostname = ""; QString username = ""; QString pass = ""; bool isConnected = theFunctions.StartSession(hostname, username, pass, 222); QVERIFY2(isConnected, "Unable to log in as "); bool isInIncoming = theFunctions.ChangeDirectory("/incoming"); QVERIFY2(isInIncoming, "Unable to change to the incoming directory"); //create a garbage file int theGarbageSize = 1024; QByteArray theGarbage(1024, 100); for (int i = 0; i < theGarbageSize; i++){ theGarbage[i] = rand()%256; } //write the garbage to a local directory QFile theGarbageFile("testgarbage.txt"); bool canOpenFile = theGarbageFile.open(QIODevice::ReadWrite | QIODevice::Text); QVERIFY2(canOpenFile, "Unable to open garbage test file"); bool wroteTestFile = theGarbageFile.write(theGarbage); QVERIFY2(wroteTestFile, "Unable to write to the garbage test file"); theGarbageFile.close(); bool wroteToTheRemote = theFunctions.PutFile(theGarbageFile.fileName(), "/home/audrey/incoming/garbage.txt"); QVERIFY2(wroteToTheRemote, "Unable to put file"); Essentially, I'm creating a garbage file, going to send it, and then retrieving it to make sure that everything's working. Logging in is done by code that looks very similar to the example code, because it's copied from the example code. The line that fails is: sftp_handle = libssh2_sftp_open_ex(sftp_session, theSendRemotePath.data(), theSendRemotePath.size(), LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC, LIBSSH2_SFTP_S_IFREG,LIBSSH2_SFTP_OPENFILE); and the login code looks like: bool SFTPSession::LogIn(const QString& inHostName, const int& inPort, const QString& inUsername, const QString& inPass) { #ifdef WIN32 WSADATA wsadata; WSAStartup(MAKEWORD(2,0), &wsadata); #endif int sock, auth_pw = 1, rc; struct sockaddr_in sin; unsigned long hostaddr; //if you don't do .data on a toAscii, then you'll get some //garbage about Ubuntu in your strings instead QByteArray theHostNameArray = inHostName.toAscii(); hostaddr = inet_addr(theHostNameArray.data()); QByteArray theUsername = inUsername.toAscii(); const char* username = theUsername.data(); QByteArray thePassword = inPass.toAscii(); const char* password = thePassword.data(); rc = libssh2_init (0); if (rc != 0) { fprintf (stderr, "libssh2 initialization failed (%d)\n", rc); return false; } /* * The application code is responsible for creating the socket * and establishing the connection */ sock = socket(AF_INET, SOCK_STREAM, 0); sin.sin_family = AF_INET; sin.sin_port = htons(inPort); sin.sin_addr.s_addr = hostaddr; if (connect(sock, (struct sockaddr*)(&sin), sizeof(struct sockaddr_in)) != 0) { fprintf(stderr, "failed to connect!\n"); return false; } /* Create a session instance */ ssh_session = libssh2_session_init(); if(!ssh_session){ fprintf(stderr, "Unable to start ssh_session"); return false; } /* ... start it up. This will trade welcome banners, exchange keys, * and setup crypto, compression, and MAC layers */ rc = libssh2_session_startup(ssh_session, sock); if(rc) { fprintf(stderr, "Failure establishing SSH session: %d\n", rc); return false; } /* At this point we havn't yet authenticated. The first thing to do * is check the hostkey's fingerprint against our known hosts Your app * may have it hard coded, may go to a file, may present it to the * user, that's your call */ const char* fingerprint = libssh2_hostkey_hash(ssh_session, LIBSSH2_HOSTKEY_HASH_SHA1); mFingerprint.fromAscii(fingerprint, 40); printf("Fingerprint: "); for(int i = 0; i < 20; i++) { fprintf(stderr, "%02X ", (unsigned char)fingerprint[i]); } printf("\n"); if (auth_pw) { /* We could authenticate via password */ if (libssh2_userauth_password(ssh_session, username, password)) { libssh2_session_disconnect(ssh_session, "Authentication by password failed\n"); libssh2_session_free(ssh_session); ssh_session = NULL; return false; } } else { /* Or by public key */ if (libssh2_userauth_publickey_fromfile(ssh_session, username, "/home/username/.ssh/id_rsa.pub", "/home/username/.ssh/id_rsa", password)) { libssh2_session_disconnect(ssh_session, "Authentication by public key failed\n"); libssh2_session_free(ssh_session); ssh_session = NULL; return false; } } fprintf(stderr, "libssh2_sftp_init()!\n"); if (ssh_session == NULL){ return false; } sftp_session = libssh2_sftp_init(ssh_session); if (!sftp_session) { libssh2_session_disconnect(ssh_session, "Unable to init SFTP session\n"); libssh2_session_free(ssh_session); return false; } /* Since we have not set non-blocking, tell libssh2 we are blocking */ libssh2_session_set_blocking(ssh_session, 1); return true; } Any thoughts here? Thanks! Mark _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 07:29:05 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p425Sirn005319; Mon, 2 May 2011 07:29:02 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p425SdsE005268 for ; Mon, 2 May 2011 07:28:41 +0200 Message-ID: <4DBE40D6.9010703@securitease.com> Date: Mon, 02 May 2011 17:27:50 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2-devel@cool.haxx.se, Mike Elston Subject: timeouts in libssh2 blocking-mode Content-Type: multipart/mixed; boundary="------------000503040209090300030305" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 07:29:05 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 07:28:42 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a multi-part message in MIME format. --------------000503040209090300030305 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hello, We've been using libssh2 for a while in some internal tests. It generally works very well, but there's an annoying problem where badly behaved ssh servers can cause lockups inside libssh2. Jan Van Boghout posted some patches to implement a simple workaround for this last year (http://www.libssh2.org/mail/libssh2-devel-archive-2010-02/0073.shtml), and I see there's a bug (http://trac.libssh2.org/ticket/160) for it as well, currently partially complete but not fully integrated (and now almost a year old) I've attached an updated patch against the tip of the master git repository which preserves the recently added keepalive support, while ensuring in the end we don't block forever. While the keepalive support works very well after the session is established, until that point, it seems that sending keepalives will cause the server to boot us off; that means we are exposed to potential endless blocking during a session creation. Since I'm new to the mailing list, I may have missed recent discussion, but the bug hasn't really moved in a year, so I thought it might be worth reinvorigating debate - comments very welcome. Regards, Matt -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality --------------000503040209090300030305 Content-Transfer-Encoding: 7bit Content-Type: text/plain; name="0001-Implement-timeout-as-per-Jan-Van-Boghout-s-suggestio.patch" Content-Disposition: attachment; filename*0="0001-Implement-timeout-as-per-Jan-Van-Boghout-s-suggestio.pa"; filename*1="tch" From 2acaa39fd4681f505b62dc4faf34c23d3d4c4ca3 Mon Sep 17 00:00:00 2001 From: Matt Lilley Date: Mon, 2 May 2011 17:21:19 +1200 Subject: [PATCH] Implement timeout as per Jan Van Boghout's suggestion in 2010 --- include/libssh2.h | 4 ++++ src/libssh2_priv.h | 3 +++ src/session.c | 43 ++++++++++++++++++++++++++++++++++++++----- 3 files changed, 45 insertions(+), 5 deletions(-) diff --git a/include/libssh2.h b/include/libssh2.h index e207411..9462360 100644 --- a/include/libssh2.h +++ b/include/libssh2.h @@ -705,6 +705,10 @@ LIBSSH2_API int libssh2_session_get_blocking(LIBSSH2_SESSION* session); LIBSSH2_API void libssh2_channel_set_blocking(LIBSSH2_CHANNEL *channel, int blocking); +LIBSSH2_API void libssh2_session_set_timeout(LIBSSH2_SESSION* session, + long timeout); +LIBSSH2_API long libssh2_session_get_timeout(LIBSSH2_SESSION* session); + /* libssh2_channel_handle_extended_data is DEPRECATED, do not use! */ LIBSSH2_API void libssh2_channel_handle_extended_data(LIBSSH2_CHANNEL *channel, int ignore_mode); diff --git a/src/libssh2_priv.h b/src/libssh2_priv.h index 231401e..5a697c7 100644 --- a/src/libssh2_priv.h +++ b/src/libssh2_priv.h @@ -559,6 +559,9 @@ struct _LIBSSH2_SESSION /* this is set to TRUE if a blocking API behavior is requested */ int api_block_mode; + /* Timeout used when blocking API behavior is active */ + long api_timeout; + /* Server's public key */ const LIBSSH2_HOSTKEY_METHOD *hostkey; void *server_hostkey_abstract; diff --git a/src/session.c b/src/session.c index e4714e8..cc5e23a 100644 --- a/src/session.c +++ b/src/session.c @@ -481,6 +481,7 @@ libssh2_session_init_ex(LIBSSH2_ALLOC_FUNC((*my_alloc)), session->free = local_free; session->realloc = local_realloc; session->abstract = abstract; + session->api_timeout = 0; /* timeout-free API by default */ session->api_block_mode = 1; /* blocking API by default */ _libssh2_debug(session, LIBSSH2_TRACE_TRANS, "New session resource allocated"); @@ -547,6 +548,7 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) int rc; int seconds_to_next; int dir; + int has_timeout; /* since libssh2 often sets EAGAIN internally before this function is called, we can decrease some amount of confusion in user programs by @@ -592,8 +594,19 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) fd_set *readfd = NULL; struct timeval tv; - tv.tv_sec = seconds_to_next; - tv.tv_usec = 0; + if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else if (session->api_timeout > 0) { + /* Seconds */ + tv.tv_sec = session->api_timeout / 1000; + /* Microseconds */ + tv.tv_usec = (session->api_timeout % 1000) * 1000; + has_timeout = 1; + } else { + has_timeout = 0; + } if(dir & LIBSSH2_SESSION_BLOCK_INBOUND) { FD_ZERO(&rfd); @@ -607,10 +620,8 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) writefd = &wfd; } - /* Note that this COULD be made to use a timeout that perhaps - could be customizable by the app or something... */ rc = select(session->socket_fd + 1, readfd, writefd, NULL, - seconds_to_next ? &tv : NULL); + has_timeout ? &tv : NULL); #endif } } @@ -1281,6 +1292,28 @@ libssh2_session_get_blocking(LIBSSH2_SESSION * session) return session->api_block_mode; } + +/* libssh2_session_set_timeout + * + * Set a session's timeout (in msec) for blocking mode, + * or 0 to disable timeouts. + */ +LIBSSH2_API void +libssh2_session_set_timeout(LIBSSH2_SESSION * session, long timeout) +{ + session->api_timeout = timeout; +} + +/* libssh2_session_get_timeout + * + * Returns a session's timeout, or 0 if disabled + */ +LIBSSH2_API long +libssh2_session_get_timeout(LIBSSH2_SESSION * session) +{ + return session->api_timeout; +} + /* * libssh2_poll_channel_read * -- 1.7.2.3.msysgit.0 --------------000503040209090300030305 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --------------000503040209090300030305-- From libssh2-devel-bounces@cool.haxx.se Mon May 2 08:33:57 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p426XY2c008422; Mon, 2 May 2011 08:33:54 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p426XXZK008416 for ; Mon, 2 May 2011 08:33:33 +0200 Received: (qmail 20481 invoked by uid 501); 2 May 2011 06:33:35 -0000 Message-ID: <20110502063335.20480.qmail@stuge.se> Date: Mon, 2 May 2011 08:33:35 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: timeouts in libssh2 blocking-mode Mail-Followup-To: libssh2-devel@cool.haxx.se References: <4DBE40D6.9010703@securitease.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <4DBE40D6.9010703@securitease.com> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 08:33:57 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 08:33:34 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Matt Lilley wrote: > badly behaved ssh servers can cause lockups inside libssh2 > http://trac.libssh2.org/ticket/160 .. > I've attached an updated patch against the tip of the master git > repository which preserves the recently added keepalive support, > while ensuring in the end we don't block forever. Nice! > Since I'm new to the mailing list, I may have missed recent > discussion, No discussion. > but the bug hasn't really moved in a year, so I thought it might be > worth reinvorigating debate - comments very welcome. I'm happy to commit this. I would like to touch up the commit message slightly however. In particular it should include some variation of "fixes #160" so that the ticket gets closed. //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 09:06:36 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4276RWC027938; Mon, 2 May 2011 09:06:35 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4276Q2Q027927; Mon, 2 May 2011 09:06:26 +0200 Date: Mon, 2 May 2011 09:06:26 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DBE40D6.9010703@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 09:06:36 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 09:06:27 +0200 (CEST) Cc: Mike Elston X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Mon, 2 May 2011, Matt Lilley wrote: > I've attached an updated patch against the tip of the master git repository I would prefer to have the timeout value to be the max time allowed for the libssh2 function to stay blocked rather than this approach which sets the maximum time for each individual select(). That will make it a lot more deterministic and understandable to users. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 09:14:49 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p427Ekr0030905; Mon, 2 May 2011 09:14:48 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p427EjVH030898 for ; Mon, 2 May 2011 09:14:45 +0200 Date: Mon, 2 May 2011 09:14:45 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Attempting to wrap sftp in Qt In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 09:14:49 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 09:14:45 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Sun, 1 May 2011, Mark Roden wrote: > I'm getting an error with trying to copy a file using sftp with > libssh2 from a windows 7 machine to a linux machine using libssh2. > The error number is 2 which error number? > and from this thread here: I failed to see how that was relevant to libssh2. Maybe I'm just dense. > The line that fails is: > > > sftp_handle = > libssh2_sftp_open_ex(sftp_session, > theSendRemotePath.data(), theSendRemotePath.size(), > LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC, > LIBSSH2_SFTP_S_IFREG,LIBSSH2_SFTP_OPENFILE); And that returns NULL and you then call libssh2_sftp_last_error() which returns 2? -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 17:58:30 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42FvuQB023132; Mon, 2 May 2011 17:58:25 +0200 Received: from mail-pw0-f54.google.com (mail-pw0-f54.google.com [209.85.160.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42Fvrmu023024 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Mon, 2 May 2011 17:57:55 +0200 Received: by pwj8 with SMTP id 8so3322770pwj.41 for ; Mon, 02 May 2011 08:57:53 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type:content-transfer-encoding; bh=YjrX/EG+Ht+MG5xNkT56f/cuJv2gi/nnZbXbcuv2MpI=; b=dqZ/AyIgZ+2ReulClwqQnScwFoElvE20avJMenYrYWdkYbYRTrDenuMY6fLpJctVpL NGShAOmG/FvhAWntZZWXQzkTUl8PvB7X/91g+U6l6MU7/gWvhz3QbW1ZUjRQnjYRlRVm qlTptMxRZRC16ZgIpDBq9FUXVeq/AXdEekw4I= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=W2MydHo8BVOisv2L33uR3gg1kprShcoHxpzc3E9vfbvpsdbRHJkEa7UjPVGhhd+iw5 E3aMLesqyrwKWOl9tFNhiQ3H4SdGHr8gah2Oc14zinfcKKlz2fv1EDA0EqTnkY188gP0 sRSXAEwyYG3hGgO0KnPInuCFMI32JZzWFk72w= MIME-Version: 1.0 Received: by 10.68.8.225 with SMTP id u1mr6280654pba.409.1304351872923; Mon, 02 May 2011 08:57:52 -0700 (PDT) Received: by 10.68.40.198 with HTTP; Mon, 2 May 2011 08:57:52 -0700 (PDT) In-Reply-To: References: Date: Mon, 2 May 2011 08:57:52 -0700 Message-ID: Subject: Re: Attempting to wrap sftp in Qt From: Mark Roden To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 17:58:30 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 17:57:55 +0200 (CEST) X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p42Fvrmu023024 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="iso-8859-1" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p42FvuQB023132 On Mon, May 2, 2011 at 12:14 AM, Daniel Stenberg wrote: > On Sun, 1 May 2011, Mark Roden wrote: > >> I'm getting an error with trying to copy a file using sftp with >> libssh2 from a windows 7 machine to a linux machine using libssh2. >> The error number is 2 > > which error number? Sorry, I don't think I was specific enough about the libssh2-ness. From debugging into the libssh2_sftp_open_ex, I get into sftp_open. From there, I get to rc = sftp_packet_requirev(sftp, 2, fopen_responses, sftp->open_request_id, &data, &data_len); (line 946 of sftp.c) Inside that function is a while statement on line 391 of sftp.c. The while statement executes once, apparently with no initial return, and then on the second go-through, the second valid response if statement is entered on line 393. Essentially, the server returns an error code, '2'. Back out into the sftp_open function, that causes the 'badness' flag to be set to 1 (line 965), presumably because that's not a good response (the comment indicates that it's a 'status' response, which is generally bad). No handle is created as a result. > >> and from this thread here: > > I failed to see how that was relevant to libssh2. Maybe I'm just dense. I don't know if it's a libssh2 error or not. I do know that I can upload files via an sftp client using the same username/password, so I suspect that I've misconfigured the connection somehow. I also know that I can get a directory listing (as per the example code), I just can't seem to be able to write using sftp. I tried using the exact code from the sftp write example, and that gave the same result. I've started experimenting with the flags, but I do believe that something's been incorrectly set up during the establishment of the connection, and was hoping for some insight there. > >> The line that fails is: >> >> >>       sftp_handle = >>           libssh2_sftp_open_ex(sftp_session, >> theSendRemotePath.data(), theSendRemotePath.size(), >> >> LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC, >> >>  LIBSSH2_SFTP_S_IFREG,LIBSSH2_SFTP_OPENFILE); > > And that returns NULL and you then call libssh2_sftp_last_error() which > returns 2? I traced the error, as described above. I haven't figured out how to see fprintf's from Qt's debugging interface in their designer. Thanks, Mark > > -- > >  / daniel.haxx.se > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 20:42:17 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42IftO0012863; Mon, 2 May 2011 20:42:14 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p42IfsX5012848 for ; Mon, 2 May 2011 20:41:55 +0200 Received: (qmail 28214 invoked by uid 501); 2 May 2011 18:42:00 -0000 Message-ID: <20110502184200.28213.qmail@stuge.se> Date: Mon, 2 May 2011 20:42:00 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: Attempting to wrap sftp in Qt Mail-Followup-To: libssh2-devel@cool.haxx.se References: MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 20:42:17 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 20:41:55 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Mark Roden wrote: > Essentially, the server returns an error code, '2'. The server sometimes also returns a text message. If this is indeed the error code from the server then it should be well defined in the SFTP spec. Likewise the error message may provide hints on what is going wrong. Both the error code and the error message are retrievable using libssh2 API after libssh2_sftp_open (or another function call) fails. > I don't know if it's a libssh2 error or not. Is the home directory for your user writable by the user? Maybe there is just a path missing somewhere to specify the upload directory. //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 21:42:32 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42JgAXd025079; Mon, 2 May 2011 21:42:29 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42Jg6ui024777 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Mon, 2 May 2011 21:42:08 +0200 Received: by pvg11 with SMTP id 11so4121137pvg.41 for ; Mon, 02 May 2011 12:42:04 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=0rDUQK2fPFf+moMnajf5wVExwT5H2WMzFB/WNhmHxkM=; b=w0Yyjgq+NHFvXOda1OBi0xa5UR31TiAIlzZph9O8ov00/fFeTPzEi9b+0fVb1Opp7l X2WHpHQ/5AUsefSg28gfVPgj54i/2VOJR+FKsRqk0NWGuy/1mi5nmCeFHweQT5ByOuTD VLl3x1d+tpi5y72IfIz7t9rOAo+Eyt5e8uuFI= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=CjWxmXROG60Xo0v3sORe6iHuEnVTpEq/pT5Va41xh3z/iihu2YaSma3UGYusYqhkOH QuCDGtR2iyqr3Cn+S30VKT5koQYiIH+KuWZuFXw0OWroSz/+fmKM5KAEHpty7egZm1yE R9idEzJLKv1s0/G7xD76eK/IF7/ZRxRa7lTrM= MIME-Version: 1.0 Received: by 10.68.8.225 with SMTP id u1mr6515738pba.409.1304365324471; Mon, 02 May 2011 12:42:04 -0700 (PDT) Received: by 10.68.40.198 with HTTP; Mon, 2 May 2011 12:42:04 -0700 (PDT) In-Reply-To: <20110502184200.28213.qmail@stuge.se> References: <20110502184200.28213.qmail@stuge.se> Date: Mon, 2 May 2011 12:42:04 -0700 Message-ID: Subject: Re: Attempting to wrap sftp in Qt From: Mark Roden To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 21:42:32 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 21:42:08 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se After further testing, it appears that it worked once, and then has stopped working. It looks like I'm not able to overwrite the file, despite setting the flags: LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC The first time, the writing works, and if I delete the file from the server using a separate session, then the writing works again. Are there other flags or things to check so that a file can be overwritten, some unix permission I'm not setting when the file is initially created? On Mon, May 2, 2011 at 11:42 AM, Peter Stuge wrote: > Mark Roden wrote: >> Essentially, the server returns an error code, '2'. > > The server sometimes also returns a text message. > > If this is indeed the error code from the server then it should be > well defined in the SFTP spec. Likewise the error message may provide > hints on what is going wrong. > > Both the error code and the error message are retrievable using > libssh2 API after libssh2_sftp_open (or another function call) fails. > > >> I don't know if it's a libssh2 error or not. > > Is the home directory for your user writable by the user? Maybe there > is just a path missing somewhere to specify the upload directory. > > > //Peter > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 22:14:02 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42KDqXG030745; Mon, 2 May 2011 22:13:59 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42KDmFN030663 for ; Mon, 2 May 2011 22:13:50 +0200 Message-ID: <4DBF104B.2010508@securitease.com> Date: Tue, 03 May 2011 08:12:59 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2-devel@cool.haxx.se, Mike Elston Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> In-Reply-To: <20110502063335.20480.qmail@stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 22:14:02 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 22:13:50 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se > I'm happy to commit this. I would like to touch up the commit message > slightly however. In particular it should include some variation of > "fixes #160" so that the ticket gets closed. > > > //Peter Ok, that sounds fine to me. Daniel has commented that he'd prefer the timeout to be the maximum time that any libssh2 function can remain blocked, which I agree would be nice, but I'm not sure how to implement - I guess each entry point could reset a counter, and then the recv() could have a 1 second timeout, which increments this counter until it reaches the specified limit, at which point the function returns with a timeout error. This is a lot of work though; every function would have to be modified (or perhaps the BLOCK_ADJUST macro?) and I'm not sure there are significant gains - after all, the keepalive code does something similar to this. All I want is a safety net to prevent our application locking up when it connects to an ssh server that we have no administrative control over. If I've missed some simple way of achieving Daniel's approach, feel free to point it out to me, and if I understand, I can have a go at implementing it. Matt -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 2 22:38:26 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42KJXGB003084; Mon, 2 May 2011 22:19:35 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42KJWYo003071; Mon, 2 May 2011 22:19:32 +0200 Date: Mon, 2 May 2011 22:19:32 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DBF104B.2010508@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 02 May 2011 22:19:36 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 02 May 2011 22:19:32 +0200 (CEST) Cc: Mike Elston X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 3 May 2011, Matt Lilley wrote: > Ok, that sounds fine to me. Daniel has commented that he'd prefer the timeout > to be the maximum time that any libssh2 function can remain blocked, which I > agree would be nice, but I'm not sure how to implement I suggest we: A) mark the time when the function is first entered. B) when select() is about to be called, we can check how long time that has elapsed since the entry time and just do the math on how long the select timeout may be (this time). -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 01:27:47 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42NRQ8Q018075; Tue, 3 May 2011 01:27:43 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42NRH3u017999 for ; Tue, 3 May 2011 01:27:23 +0200 Message-ID: <4DBF3D9D.7080800@securitease.com> Date: Tue, 03 May 2011 11:26:21 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2 development , Mike Elston Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> In-Reply-To: Content-Type: multipart/mixed; boundary="------------020105070508070107040902" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 01:27:47 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 01:27:24 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a multi-part message in MIME format. --------------020105070508070107040902 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed > I suggest we: > > A) mark the time when the function is first entered. > > B) when select() is about to be called, we can check how long time that > has elapsed since the entry time and just do the math on how long the > select timeout may be (this time). Ok, that seems reasonable to me. I've attached a new patch that does this (I think), and still passes my test cases. I assumed that because the macro BLOCK_ADJUST (and BLOCK_ADJUST_ERRNO) were the only places we'd end up in _libssh2_wait_socket, that this was as good a place as any to mark the entry time, and they all seem to be at the beginning of exposed API functions. Any further thoughts? Matt > > -- > > / daniel.haxx.se > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality --------------020105070508070107040902 Content-Transfer-Encoding: 7bit Content-Type: text/plain; name="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" Content-Disposition: attachment; filename="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" From dae57c891dd99cff7b8811b05a39593f9237ba55 Mon Sep 17 00:00:00 2001 From: Matt Lilley Date: Tue, 3 May 2011 11:20:24 +1200 Subject: [PATCH] Fixes bug #160 as per Daniel's suggestion --- include/libssh2.h | 4 ++++ src/libssh2_priv.h | 3 +++ src/session.c | 51 +++++++++++++++++++++++++++++++++++++++++++++------ src/session.h | 19 ++++++++++++------- 4 files changed, 64 insertions(+), 13 deletions(-) diff --git a/include/libssh2.h b/include/libssh2.h index e207411..9462360 100644 --- a/include/libssh2.h +++ b/include/libssh2.h @@ -705,6 +705,10 @@ LIBSSH2_API int libssh2_session_get_blocking(LIBSSH2_SESSION* session); LIBSSH2_API void libssh2_channel_set_blocking(LIBSSH2_CHANNEL *channel, int blocking); +LIBSSH2_API void libssh2_session_set_timeout(LIBSSH2_SESSION* session, + long timeout); +LIBSSH2_API long libssh2_session_get_timeout(LIBSSH2_SESSION* session); + /* libssh2_channel_handle_extended_data is DEPRECATED, do not use! */ LIBSSH2_API void libssh2_channel_handle_extended_data(LIBSSH2_CHANNEL *channel, int ignore_mode); diff --git a/src/libssh2_priv.h b/src/libssh2_priv.h index 231401e..5a697c7 100644 --- a/src/libssh2_priv.h +++ b/src/libssh2_priv.h @@ -559,6 +559,9 @@ struct _LIBSSH2_SESSION /* this is set to TRUE if a blocking API behavior is requested */ int api_block_mode; + /* Timeout used when blocking API behavior is active */ + long api_timeout; + /* Server's public key */ const LIBSSH2_HOSTKEY_METHOD *hostkey; void *server_hostkey_abstract; diff --git a/src/session.c b/src/session.c index e4714e8..edf98d8 100644 --- a/src/session.c +++ b/src/session.c @@ -481,6 +481,7 @@ libssh2_session_init_ex(LIBSSH2_ALLOC_FUNC((*my_alloc)), session->free = local_free; session->realloc = local_realloc; session->abstract = abstract; + session->api_timeout = 0; /* timeout-free API by default */ session->api_block_mode = 1; /* blocking API by default */ _libssh2_debug(session, LIBSSH2_TRACE_TRANS, "New session resource allocated"); @@ -542,11 +543,12 @@ libssh2_session_callback_set(LIBSSH2_SESSION * session, * Utility function that waits for action on the socket. Returns 0 when ready * to run again or error on timeout. */ -int _libssh2_wait_socket(LIBSSH2_SESSION *session) +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t start_time) { int rc; int seconds_to_next; int dir; + int has_timeout; /* since libssh2 often sets EAGAIN internally before this function is called, we can decrease some amount of confusion in user programs by @@ -592,8 +594,25 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) fd_set *readfd = NULL; struct timeval tv; - tv.tv_sec = seconds_to_next; - tv.tv_usec = 0; + if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else if (session->api_timeout > 0) { + time_t now = time (NULL); + long elapsed_ms = (long)(1000*difftime(start_time, now)); + if (elapsed_ms > session->api_timeout) { + session->err_code = LIBSSH2_ERROR_TIMEOUT; + return LIBSSH2_ERROR_TIMEOUT; + } + /* Seconds */ + tv.tv_sec = (session->api_timeout - elapsed_ms) / 1000; + /* Microseconds */ + tv.tv_usec = ((session->api_timeout - elapsed_ms) % 1000) * 1000; + has_timeout = 1; + } else { + has_timeout = 0; + } if(dir & LIBSSH2_SESSION_BLOCK_INBOUND) { FD_ZERO(&rfd); @@ -607,10 +626,8 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) writefd = &wfd; } - /* Note that this COULD be made to use a timeout that perhaps - could be customizable by the app or something... */ rc = select(session->socket_fd + 1, readfd, writefd, NULL, - seconds_to_next ? &tv : NULL); + has_timeout ? &tv : NULL); #endif } } @@ -1281,6 +1298,28 @@ libssh2_session_get_blocking(LIBSSH2_SESSION * session) return session->api_block_mode; } + +/* libssh2_session_set_timeout + * + * Set a session's timeout (in msec) for blocking mode, + * or 0 to disable timeouts. + */ +LIBSSH2_API void +libssh2_session_set_timeout(LIBSSH2_SESSION * session, long timeout) +{ + session->api_timeout = timeout; +} + +/* libssh2_session_get_timeout + * + * Returns a session's timeout, or 0 if disabled + */ +LIBSSH2_API long +libssh2_session_get_timeout(LIBSSH2_SESSION * session) +{ + return session->api_timeout; +} + /* * libssh2_poll_channel_read * diff --git a/src/session.h b/src/session.h index 4eac3a6..c7e01e9 100644 --- a/src/session.h +++ b/src/session.h @@ -51,17 +51,20 @@ function. */ -#define BLOCK_ADJUST(rc,sess,x) \ +#define BLOCK_ADJUST(rc,sess,x) { \ + time_t entry_time = time (NULL); \ do { \ rc = x; \ /* the order of the check below is important to properly deal with the case when the 'sess' is freed */ \ if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ break; \ - rc = _libssh2_wait_socket(sess); \ + rc = _libssh2_wait_socket(sess, entry_time); \ if(rc) \ break; \ - } while(1) + } while(1); \ +} + /* * For functions that returns a pointer, we need to check if the API is @@ -69,7 +72,8 @@ * immediately. If the API is blocking and we get a NULL we check the errno * and *only* if that is EAGAIN we loop and wait for socket action. */ -#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ +#define BLOCK_ADJUST_ERRNO(ptr,sess,x) { \ + time_t entry_time = time (NULL); \ do { \ int rc; \ ptr = x; \ @@ -77,13 +81,14 @@ (ptr != NULL) || \ (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ break; \ - rc = _libssh2_wait_socket(sess); \ + rc = _libssh2_wait_socket(sess, entry_time); \ if(rc) \ break; \ - } while(1) + } while(1); \ +} -int _libssh2_wait_socket(LIBSSH2_SESSION *session); +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t entry_time); /* this is the lib-internal set blocking function */ int _libssh2_session_set_blocking(LIBSSH2_SESSION * session, int blocking); -- 1.7.2.3.msysgit.0 --------------020105070508070107040902 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --------------020105070508070107040902-- From libssh2-devel-bounces@cool.haxx.se Tue May 3 01:59:40 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p42NxVoX003230; Tue, 3 May 2011 01:59:39 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p42NxTw9003227 for ; Tue, 3 May 2011 01:59:29 +0200 Received: (qmail 32307 invoked by uid 501); 2 May 2011 23:59:37 -0000 Message-ID: <20110502235937.32306.qmail@stuge.se> Date: Tue, 3 May 2011 01:59:37 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: timeouts in libssh2 blocking-mode Mail-Followup-To: libssh2-devel@cool.haxx.se References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <4DBF3D9D.7080800@securitease.com> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 01:59:40 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 01:59:29 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Matt Lilley wrote: > Ok, that seems reasonable to me. I've attached a new patch that does this > (I think), and still passes my test cases. I'll let you and Daniel sort the implementation details. Just one comment: > +++ b/src/session.h > @@ -51,17 +51,20 @@ > function. > > */ > -#define BLOCK_ADJUST(rc,sess,x) \ > +#define BLOCK_ADJUST(rc,sess,x) { \ > + time_t entry_time = time (NULL); \ > do { \ > rc = x; \ > /* the order of the check below is important to properly deal with the > case when the 'sess' is freed */ \ > if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ > break; \ > - rc = _libssh2_wait_socket(sess); \ > + rc = _libssh2_wait_socket(sess, entry_time); \ > if(rc) \ > break; \ > - } while(1) > + } while(1); \ > +} > + No. Using do {} while() must not change (you add braces) since it allows the macro to successfully be used as a function in all circumstances in C code. In this case you would add an outer do {} while() as such: #define BLOCK_ADJUST(rc,sess,x) { \ do { \ time_t entry_time ... \ do { \ ... } while(1); } while(0) Also, please make sure to maintain code style (whitespace changes) in the code that you work on. Thanks! Btw, it doesn't make sense to have while(1) immediately preceded by a conditional break IMO. Might as well fix that too while touching this: .. while(!rc); > @@ -69,7 +72,8 @@ > * immediately. If the API is blocking and we get a NULL we check the errno > * and *only* if that is EAGAIN we loop and wait for socket action. > */ > -#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ > +#define BLOCK_ADJUST_ERRNO(ptr,sess,x) { \ Same comment as above. //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 02:20:17 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p430KABo017140; Tue, 3 May 2011 02:20:16 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p430K5Zh016435 for ; Tue, 3 May 2011 02:20:08 +0200 Message-ID: <4DBF4A05.9050403@securitease.com> Date: Tue, 03 May 2011 12:19:17 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2 development , Mike Elston Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> In-Reply-To: <20110502235937.32306.qmail@stuge.se> Content-Type: multipart/mixed; boundary="------------060502080100020503010701" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 02:20:17 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 02:20:08 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a multi-part message in MIME format. --------------060502080100020503010701 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed > No. Using do {} while() must not change (you add braces) since it > allows the macro to successfully be used as a function in all > circumstances in C code. In this case you would add an outer > do {} while() as such: Ok, that's fair enough. Changed > Also, please make sure to maintain code style (whitespace changes) in > the code that you work on. Thanks! I assume you're talking about the spaces between the backslash and the code in the macro? If so, I think I've fixed that now (I didn't notice my editor had done it). If it's something else, can you be more specific so I can fix that as well? > Btw, it doesn't make sense to have while(1) immediately preceded by a > conditional break IMO. Agreed, changed. Matt -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality --------------060502080100020503010701 Content-Transfer-Encoding: 7bit Content-Type: text/plain; name="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" Content-Disposition: attachment; filename="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" From 1af4d98907ac318e0cee8675306a05e4d223de65 Mon Sep 17 00:00:00 2001 From: Matt Lilley Date: Tue, 3 May 2011 11:20:24 +1200 Subject: [PATCH] Fixes bug #160 as per Daniel's suggestion --- include/libssh2.h | 4 ++++ src/libssh2_priv.h | 3 +++ src/session.c | 51 +++++++++++++++++++++++++++++++++++++++++++++------ src/session.h | 44 +++++++++++++++++++++++--------------------- 4 files changed, 75 insertions(+), 27 deletions(-) diff --git a/include/libssh2.h b/include/libssh2.h index e207411..9462360 100644 --- a/include/libssh2.h +++ b/include/libssh2.h @@ -705,6 +705,10 @@ LIBSSH2_API int libssh2_session_get_blocking(LIBSSH2_SESSION* session); LIBSSH2_API void libssh2_channel_set_blocking(LIBSSH2_CHANNEL *channel, int blocking); +LIBSSH2_API void libssh2_session_set_timeout(LIBSSH2_SESSION* session, + long timeout); +LIBSSH2_API long libssh2_session_get_timeout(LIBSSH2_SESSION* session); + /* libssh2_channel_handle_extended_data is DEPRECATED, do not use! */ LIBSSH2_API void libssh2_channel_handle_extended_data(LIBSSH2_CHANNEL *channel, int ignore_mode); diff --git a/src/libssh2_priv.h b/src/libssh2_priv.h index 231401e..5a697c7 100644 --- a/src/libssh2_priv.h +++ b/src/libssh2_priv.h @@ -559,6 +559,9 @@ struct _LIBSSH2_SESSION /* this is set to TRUE if a blocking API behavior is requested */ int api_block_mode; + /* Timeout used when blocking API behavior is active */ + long api_timeout; + /* Server's public key */ const LIBSSH2_HOSTKEY_METHOD *hostkey; void *server_hostkey_abstract; diff --git a/src/session.c b/src/session.c index e4714e8..edf98d8 100644 --- a/src/session.c +++ b/src/session.c @@ -481,6 +481,7 @@ libssh2_session_init_ex(LIBSSH2_ALLOC_FUNC((*my_alloc)), session->free = local_free; session->realloc = local_realloc; session->abstract = abstract; + session->api_timeout = 0; /* timeout-free API by default */ session->api_block_mode = 1; /* blocking API by default */ _libssh2_debug(session, LIBSSH2_TRACE_TRANS, "New session resource allocated"); @@ -542,11 +543,12 @@ libssh2_session_callback_set(LIBSSH2_SESSION * session, * Utility function that waits for action on the socket. Returns 0 when ready * to run again or error on timeout. */ -int _libssh2_wait_socket(LIBSSH2_SESSION *session) +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t start_time) { int rc; int seconds_to_next; int dir; + int has_timeout; /* since libssh2 often sets EAGAIN internally before this function is called, we can decrease some amount of confusion in user programs by @@ -592,8 +594,25 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) fd_set *readfd = NULL; struct timeval tv; - tv.tv_sec = seconds_to_next; - tv.tv_usec = 0; + if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else if (session->api_timeout > 0) { + time_t now = time (NULL); + long elapsed_ms = (long)(1000*difftime(start_time, now)); + if (elapsed_ms > session->api_timeout) { + session->err_code = LIBSSH2_ERROR_TIMEOUT; + return LIBSSH2_ERROR_TIMEOUT; + } + /* Seconds */ + tv.tv_sec = (session->api_timeout - elapsed_ms) / 1000; + /* Microseconds */ + tv.tv_usec = ((session->api_timeout - elapsed_ms) % 1000) * 1000; + has_timeout = 1; + } else { + has_timeout = 0; + } if(dir & LIBSSH2_SESSION_BLOCK_INBOUND) { FD_ZERO(&rfd); @@ -607,10 +626,8 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) writefd = &wfd; } - /* Note that this COULD be made to use a timeout that perhaps - could be customizable by the app or something... */ rc = select(session->socket_fd + 1, readfd, writefd, NULL, - seconds_to_next ? &tv : NULL); + has_timeout ? &tv : NULL); #endif } } @@ -1281,6 +1298,28 @@ libssh2_session_get_blocking(LIBSSH2_SESSION * session) return session->api_block_mode; } + +/* libssh2_session_set_timeout + * + * Set a session's timeout (in msec) for blocking mode, + * or 0 to disable timeouts. + */ +LIBSSH2_API void +libssh2_session_set_timeout(LIBSSH2_SESSION * session, long timeout) +{ + session->api_timeout = timeout; +} + +/* libssh2_session_get_timeout + * + * Returns a session's timeout, or 0 if disabled + */ +LIBSSH2_API long +libssh2_session_get_timeout(LIBSSH2_SESSION * session) +{ + return session->api_timeout; +} + /* * libssh2_poll_channel_read * diff --git a/src/session.h b/src/session.h index 4eac3a6..aa022fb 100644 --- a/src/session.h +++ b/src/session.h @@ -53,15 +53,16 @@ */ #define BLOCK_ADJUST(rc,sess,x) \ do { \ - rc = x; \ - /* the order of the check below is important to properly deal with the - case when the 'sess' is freed */ \ - if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + rc = x; \ + /* the order of the check below is important to properly deal with the + case when the 'sess' is freed */ \ + if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(rc); \ + } while(0) /* * For functions that returns a pointer, we need to check if the API is @@ -69,21 +70,22 @@ * immediately. If the API is blocking and we get a NULL we check the errno * and *only* if that is EAGAIN we loop and wait for socket action. */ -#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ +#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ do { \ - int rc; \ - ptr = x; \ - if(!sess->api_block_mode || \ - (ptr != NULL) || \ - (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + int rc; \ + ptr = x; \ + if(!sess->api_block_mode || \ + (ptr != NULL) || \ + (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(!rc); \ + } while(0) -int _libssh2_wait_socket(LIBSSH2_SESSION *session); +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t entry_time); /* this is the lib-internal set blocking function */ int _libssh2_session_set_blocking(LIBSSH2_SESSION * session, int blocking); -- 1.7.2.3.msysgit.0 --------------060502080100020503010701 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --------------060502080100020503010701-- From libssh2-devel-bounces@cool.haxx.se Tue May 3 02:48:33 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p430mIxd004241; Tue, 3 May 2011 02:48:32 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p430mHhl004228 for ; Tue, 3 May 2011 02:48:17 +0200 Received: (qmail 5080 invoked by uid 501); 3 May 2011 00:48:25 -0000 Message-ID: <20110503004825.5079.qmail@stuge.se> Date: Tue, 3 May 2011 02:48:25 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: timeouts in libssh2 blocking-mode Mail-Followup-To: libssh2-devel@cool.haxx.se References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <4DBF4A05.9050403@securitease.com> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 02:48:33 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 02:48:18 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Matt Lilley wrote: >> Btw, it doesn't make sense to have while(1) immediately preceded by a >> conditional break IMO. > Agreed, changed. .. > +++ b/src/session.h > @@ -53,15 +53,16 @@ .. > - if(rc) \ > - break; \ > - } while(1) .. > + } while(rc); \ I think you inverted the logic. Add the ! I wrote. :) //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 03:03:13 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43137xA018555; Tue, 3 May 2011 03:03:12 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43130qo018141 for ; Tue, 3 May 2011 03:03:04 +0200 Message-ID: <4DBF5418.4040706@securitease.com> Date: Tue, 03 May 2011 13:02:16 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> In-Reply-To: <20110503004825.5079.qmail@stuge.se> Content-Type: multipart/mixed; boundary="------------050005080900000902000907" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 03:03:13 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 03:03:05 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a multi-part message in MIME format. --------------050005080900000902000907 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Wow. I need more coffee, I think. I got one of the two, at least. 50% isn't bad? > Matt Lilley wrote: >>> Btw, it doesn't make sense to have while(1) immediately preceded by a >>> conditional break IMO. >> Agreed, changed. > .. >> +++ b/src/session.h >> @@ -53,15 +53,16 @@ > .. >> - if(rc) \ >> - break; \ >> - } while(1) > .. >> + } while(rc); \ > I think you inverted the logic. Add the ! I wrote. :) > > > //Peter > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality --------------050005080900000902000907 Content-Transfer-Encoding: 7bit Content-Type: text/plain; name="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" Content-Disposition: attachment; filename="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" From 9b6453b373f40e49ada7696c39a9fa9f777fc48a Mon Sep 17 00:00:00 2001 From: Matt Lilley Date: Tue, 3 May 2011 11:20:24 +1200 Subject: [PATCH] Fixes bug #160 as per Daniel's suggestion --- include/libssh2.h | 4 ++++ src/libssh2_priv.h | 3 +++ src/session.c | 51 +++++++++++++++++++++++++++++++++++++++++++++------ src/session.h | 44 +++++++++++++++++++++++--------------------- 4 files changed, 75 insertions(+), 27 deletions(-) diff --git a/include/libssh2.h b/include/libssh2.h index e207411..9462360 100644 --- a/include/libssh2.h +++ b/include/libssh2.h @@ -705,6 +705,10 @@ LIBSSH2_API int libssh2_session_get_blocking(LIBSSH2_SESSION* session); LIBSSH2_API void libssh2_channel_set_blocking(LIBSSH2_CHANNEL *channel, int blocking); +LIBSSH2_API void libssh2_session_set_timeout(LIBSSH2_SESSION* session, + long timeout); +LIBSSH2_API long libssh2_session_get_timeout(LIBSSH2_SESSION* session); + /* libssh2_channel_handle_extended_data is DEPRECATED, do not use! */ LIBSSH2_API void libssh2_channel_handle_extended_data(LIBSSH2_CHANNEL *channel, int ignore_mode); diff --git a/src/libssh2_priv.h b/src/libssh2_priv.h index 231401e..5a697c7 100644 --- a/src/libssh2_priv.h +++ b/src/libssh2_priv.h @@ -559,6 +559,9 @@ struct _LIBSSH2_SESSION /* this is set to TRUE if a blocking API behavior is requested */ int api_block_mode; + /* Timeout used when blocking API behavior is active */ + long api_timeout; + /* Server's public key */ const LIBSSH2_HOSTKEY_METHOD *hostkey; void *server_hostkey_abstract; diff --git a/src/session.c b/src/session.c index e4714e8..edf98d8 100644 --- a/src/session.c +++ b/src/session.c @@ -481,6 +481,7 @@ libssh2_session_init_ex(LIBSSH2_ALLOC_FUNC((*my_alloc)), session->free = local_free; session->realloc = local_realloc; session->abstract = abstract; + session->api_timeout = 0; /* timeout-free API by default */ session->api_block_mode = 1; /* blocking API by default */ _libssh2_debug(session, LIBSSH2_TRACE_TRANS, "New session resource allocated"); @@ -542,11 +543,12 @@ libssh2_session_callback_set(LIBSSH2_SESSION * session, * Utility function that waits for action on the socket. Returns 0 when ready * to run again or error on timeout. */ -int _libssh2_wait_socket(LIBSSH2_SESSION *session) +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t start_time) { int rc; int seconds_to_next; int dir; + int has_timeout; /* since libssh2 often sets EAGAIN internally before this function is called, we can decrease some amount of confusion in user programs by @@ -592,8 +594,25 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) fd_set *readfd = NULL; struct timeval tv; - tv.tv_sec = seconds_to_next; - tv.tv_usec = 0; + if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else if (session->api_timeout > 0) { + time_t now = time (NULL); + long elapsed_ms = (long)(1000*difftime(start_time, now)); + if (elapsed_ms > session->api_timeout) { + session->err_code = LIBSSH2_ERROR_TIMEOUT; + return LIBSSH2_ERROR_TIMEOUT; + } + /* Seconds */ + tv.tv_sec = (session->api_timeout - elapsed_ms) / 1000; + /* Microseconds */ + tv.tv_usec = ((session->api_timeout - elapsed_ms) % 1000) * 1000; + has_timeout = 1; + } else { + has_timeout = 0; + } if(dir & LIBSSH2_SESSION_BLOCK_INBOUND) { FD_ZERO(&rfd); @@ -607,10 +626,8 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) writefd = &wfd; } - /* Note that this COULD be made to use a timeout that perhaps - could be customizable by the app or something... */ rc = select(session->socket_fd + 1, readfd, writefd, NULL, - seconds_to_next ? &tv : NULL); + has_timeout ? &tv : NULL); #endif } } @@ -1281,6 +1298,28 @@ libssh2_session_get_blocking(LIBSSH2_SESSION * session) return session->api_block_mode; } + +/* libssh2_session_set_timeout + * + * Set a session's timeout (in msec) for blocking mode, + * or 0 to disable timeouts. + */ +LIBSSH2_API void +libssh2_session_set_timeout(LIBSSH2_SESSION * session, long timeout) +{ + session->api_timeout = timeout; +} + +/* libssh2_session_get_timeout + * + * Returns a session's timeout, or 0 if disabled + */ +LIBSSH2_API long +libssh2_session_get_timeout(LIBSSH2_SESSION * session) +{ + return session->api_timeout; +} + /* * libssh2_poll_channel_read * diff --git a/src/session.h b/src/session.h index 4eac3a6..2866fcb 100644 --- a/src/session.h +++ b/src/session.h @@ -53,15 +53,16 @@ */ #define BLOCK_ADJUST(rc,sess,x) \ do { \ - rc = x; \ - /* the order of the check below is important to properly deal with the - case when the 'sess' is freed */ \ - if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + rc = x; \ + /* the order of the check below is important to properly deal with the + case when the 'sess' is freed */ \ + if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(!rc); \ + } while(0) /* * For functions that returns a pointer, we need to check if the API is @@ -69,21 +70,22 @@ * immediately. If the API is blocking and we get a NULL we check the errno * and *only* if that is EAGAIN we loop and wait for socket action. */ -#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ +#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ do { \ - int rc; \ - ptr = x; \ - if(!sess->api_block_mode || \ - (ptr != NULL) || \ - (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + int rc; \ + ptr = x; \ + if(!sess->api_block_mode || \ + (ptr != NULL) || \ + (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(!rc); \ + } while(0) -int _libssh2_wait_socket(LIBSSH2_SESSION *session); +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t entry_time); /* this is the lib-internal set blocking function */ int _libssh2_session_set_blocking(LIBSSH2_SESSION * session, int blocking); -- 1.7.2.3.msysgit.0 --------------050005080900000902000907 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --------------050005080900000902000907-- From libssh2-devel-bounces@cool.haxx.se Tue May 3 14:24:44 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43COPSJ011855; Tue, 3 May 2011 14:24:41 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43CONMh011850 for ; Tue, 3 May 2011 14:24:23 +0200 Date: Tue, 3 May 2011 14:24:23 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DBF5418.4040706@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 14:24:44 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 14:24:23 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 3 May 2011, Matt Lilley wrote: I like it! I only have this little thought. The patch has this section: + if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else if (session->api_timeout > 0) { + time_t now = time (NULL); + long elapsed_ms = (long)(1000*difftime(start_time, now)); ... which then basically makes it _not_ obey the timeout if there's a keep-alive timeout in action, which I'm not sure is desirable. Just because there's a possible keep-alive timer firing off in the future doesn't mean your timeout is no longer valid. I think the maximum time allowed to select() should also affect the timeout even if seconds_to_next is non-zero. Thoughts? -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 14:27:21 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43CRJqp014942; Tue, 3 May 2011 14:27:20 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43CRI3P014935 for ; Tue, 3 May 2011 14:27:18 +0200 Date: Tue, 3 May 2011 14:27:18 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Attempting to wrap sftp in Qt In-Reply-To: Message-ID: References: <20110502184200.28213.qmail@stuge.se> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 14:27:21 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 14:27:18 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Mon, 2 May 2011, Mark Roden wrote: > It looks like I'm not able to overwrite the file, despite setting the flags: > LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC > > The first time, the writing works, and if I delete the file from the server > using a separate session, then the writing works again. > > Are there other flags or things to check so that a file can be overwritten, > some unix permission I'm not setting when the file is initially created? If the server refuses the operation then so it does, there's not a lot libssh2 can do about to convince that it should do something else. If there would be permission problems then surely you shouldn't be allowed to remove it either? I would look into the server logs for clues. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 15:22:37 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43DMMPX017918; Tue, 3 May 2011 15:22:35 +0200 Received: from mail-pw0-f54.google.com (mail-pw0-f54.google.com [209.85.160.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43DMIAJ017877 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 3 May 2011 15:22:20 +0200 Received: by pwj8 with SMTP id 8so53089pwj.41 for ; Tue, 03 May 2011 06:22:17 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type:content-transfer-encoding; bh=MOyzC22WeDKbpo1oDWkHWZ+JPs00wZrjzPvM7LLRUWI=; b=BdpU5b9/B8WN5qMMHGQGJMwmtoemTRmEFZTNV08OJnYhZUooIxOqOf/QAiejLtDnM5 cPOpEWo7Pu9aZBGo5pz6sAzHYxrY9kkc/1yH1BHUruURzN2vbqK1OkNxY5bocQ5jGkUR FRURhhyw+sXjDNlFoWZVfjb6pH4EnniKZ6mJM= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=p5dcqiQ7RepbnfWwc1QVsx/l/gS8o1baK6V+2am6oiVeR1AjDx7L6vCzfR2wiWBtlJ hINj8LwIG1zoTWLXe5pUVTbKJytfHtB7zZ18q/zGGjDfz8VYMOBwiFyTExMAiXqvoIMj bhRylMUcFVf7151P2FCAHaM1xRy6fB7ziQCBo= MIME-Version: 1.0 Received: by 10.68.8.225 with SMTP id u1mr7734421pba.409.1304428937798; Tue, 03 May 2011 06:22:17 -0700 (PDT) Received: by 10.68.40.198 with HTTP; Tue, 3 May 2011 06:22:17 -0700 (PDT) In-Reply-To: References: <20110502184200.28213.qmail@stuge.se> Date: Tue, 3 May 2011 06:22:17 -0700 Message-ID: Subject: Re: Attempting to wrap sftp in Qt From: Mark Roden To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 15:22:37 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 15:22:20 +0200 (CEST) X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p43DMIAJ017877 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="iso-8859-1" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p43DMMPX017918 Yeah, you're right. This is a server-side issue. I thought I was establishing the connections incorrectly, and I was. I needed to be more careful about the way that I handled the handles, as well as making sure that I was going to the right path. It's now working, and all the tests are working as welll. Thanks for the feedback, Mark On Tue, May 3, 2011 at 5:27 AM, Daniel Stenberg wrote: > On Mon, 2 May 2011, Mark Roden wrote: > >> It looks like I'm not able to overwrite the file, despite setting the >> flags: LIBSSH2_FXF_WRITE|LIBSSH2_FXF_CREAT|LIBSSH2_FXF_TRUNC >> >> The first time, the writing works, and if I delete the file from the >> server using a separate session, then the writing works again. >> >> Are there other flags or things to check so that a file can be >> overwritten, some unix permission I'm not setting when the file is initially >> created? > > If the server refuses the operation then so it does, there's not a lot > libssh2 can do about to convince that it should do something else. If there > would be permission problems then surely you shouldn't be allowed to remove > it either? > > I would look into the server logs for clues. > > -- > >  / daniel.haxx.se > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 22:57:14 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43KuuD7022101; Tue, 3 May 2011 22:57:11 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43Kureo021992 for ; Tue, 3 May 2011 22:56:55 +0200 Message-ID: <4DC06BDA.2040605@securitease.com> Date: Wed, 04 May 2011 08:55:54 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> In-Reply-To: X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 22:57:14 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 22:56:56 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se > ... which then basically makes it _not_ obey the timeout if there's a > keep-alive timeout in action, which I'm not sure is desirable. Just > because there's a possible keep-alive timer firing off in the future > doesn't mean your timeout is no longer valid. > > I think the maximum time allowed to select() should also affect the > timeout even if seconds_to_next is non-zero. > > Thoughts? Well, for my puposes, the timeout is a safety-net. I don't really care when it triggers the call to exit with an error, so long as it does /at some point/ (and that this point can be configured to be on the order of seconds or minutes, rather than hours or days!). This is why my original approach was on a per-select rather than per-call basis - I only really cared about preventing endless blocking, not maintaining a guaranteed maximum time inside the library. If you'd prefer the api_timeout to have precedence over the keep-alive, that's fine by me, or we could take the minimum of the two if you want. I'm happy with either; if you have a preference, let me know and I can modify the patch, or feel free to make the change yourself as you see fit. Matt -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 23:10:15 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43LA5Fc002324; Tue, 3 May 2011 23:10:13 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43LA3WB002133 for ; Tue, 3 May 2011 23:10:03 +0200 Date: Tue, 3 May 2011 23:10:03 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DC06BDA.2040605@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> <4DC06BDA.2040605@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 23:10:14 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 23:10:03 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Wed, 4 May 2011, Matt Lilley wrote: > Well, for my puposes, the timeout is a safety-net. I don't really care when > it triggers the call to exit with an error, so long as it does /at some > point/ (and that this point can be configured to be on the order of seconds > or minutes, rather than hours or days!). This is why my original approach > was on a per-select rather than per-call basis - I only really cared about > preventing endless blocking, not maintaining a guaranteed maximum time > inside the library. I figured that, and I realize I'm arguing about a really odd edge-case that probably won't ever happen to most users. I'm just striving to make the API as clear as possible and offer as few surprises as possible to the user of it. It also makes it easier to document and describe. > If you'd prefer the api_timeout to have precedence over the keep-alive, > that's fine by me, or we could take the minimum of the two if you want. I'm > happy with either; if you have a preference, let me know and I can modify > the patch, or feel free to make the change yourself as you see fit. I'd like that if the user-set timeout is less than the keep-alive timer has left, it should use the api_timeout and return an error if it hits that timeout. If you update the patch, I'll offer to write the first draft of the man pages for the two new functions! -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 3 23:52:22 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43LqDdZ014633; Tue, 3 May 2011 23:52:20 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43Lq9do014510 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 3 May 2011 23:52:11 +0200 Received: by pvg11 with SMTP id 11so293568pvg.41 for ; Tue, 03 May 2011 14:52:11 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:date:message-id:subject:from:to :content-type; bh=G4+PL376+Xp7CpD5BShTJQ3W+A0GQQYRxTL2OVMdMPk=; b=SQ3et1zCmomM/cNEyUONx4qCRJ6hWq3XyVt9mEussqIWidEwjRXbSyRd7nyJuuZG7O E5AL50N9g0V1VgVhpgTdc6semXS2IfKtRKaWCcLrEyDwVC7F99sPVp/Um0cjVOqZ02Z1 iG45GdnNJbY/8031cDT+qaEyTSE5QAOKLF5ig= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=iLrmsb8YBl56DFWKuVXf8ErPMq6buS6EJfEr6WEWYf+kxyigbbS4BZo3+PhcMRwbR9 U1kLlk42KDA4LlmR7oW04eMPsKRCoPQGCrV6sg3XflRABZnvfb1v4EO5O+Cshd0rTVcr nxJxeB8a93t1TCJxw5zen4eEVJyG8lcokJxys= MIME-Version: 1.0 Received: by 10.142.128.9 with SMTP id a9mr183357wfd.173.1304459531508; Tue, 03 May 2011 14:52:11 -0700 (PDT) Received: by 10.142.212.16 with HTTP; Tue, 3 May 2011 14:52:11 -0700 (PDT) Date: Tue, 3 May 2011 17:52:11 -0400 Message-ID: Subject: libssh2_channel_exec guidance From: bayan To: libssh2-devel@cool.haxx.se X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 03 May 2011 23:52:22 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 03 May 2011 23:52:12 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1144358550==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1144358550== Content-Type: multipart/alternative; boundary=000e0cd32fd2ffdbe304a2662690 --000e0cd32fd2ffdbe304a2662690 Content-Type: text/plain; charset=ISO-8859-1 Hi all, I am new here, and would like to thank the people who work hard to provide this awesome library for us!! I have a question that I have been unable to figure out, after searching around and not coming up with anything, I was hoping to get guidance from the hive-mind. I seem to be having an issue using libssh2_channel_exec when sending X amount of data across the channel. To be more specific, the code looks like this: ssh_channel = libssh2_channel_open_session(ssh_session); libssh2_channel_exec(m_ssh_channel, command); "command" here, will simply execute something that pumps data to the location I want it. Everything is working all fine and dandy (oh ya, dandy!), up until I transfer X amount of (large?) data. The amount of data transferred is exactly the same every time, at which point it completely halts all transfer. It does take a while - it is a decent size of information. What I am wondering is if this is something that is known, or if I am doing something incorrectly. Alternatively to test this, what I have done is fire up Putty, and use that to manually send the "command", to see what happens. It is able to execute correctly, seemingly infinitely. What I would like, is for the data stream to not just stop at X amount of data transferred, I would like it to continue infinitely similar to how it works when I manually send the command using Putty. Please let me know if there is anything else I can provide for more help! Cheers! Ben --000e0cd32fd2ffdbe304a2662690 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
Hi all,

I am new here, and would like to than= k the people who work hard to provide this awesome library for us!!

= I have a question that I have been unable to figure out, after searching ar= ound and not coming up with anything, I was hoping to get guidance from the= =A0hive-mind.

I seem to be having an issue using libssh2_channel_exec= when sending X amount of data across the channel.

To be more specific, the code looks like this:

ssh_channel =3D libssh2_channel_open_session(ssh_session);
libssh= 2_channel_exec(m_ssh_channel, command);

"comm= and" here, will simply execute something that pumps data to the locati= on I want it.

Everything is working all fine and dandy (oh ya, dandy!= ), up until I transfer X amount of (large?) data. =A0The amount of data tra= nsferred is exactly the same=A0every time, at which point it completely hal= ts all transfer.

It does take a while - it is a decent size of information. =A0What I am= wondering is if this is something that is known, or if I am doing somethin= g incorrectly.

Alternatively=A0to test this, what I have = done is fire up Putty, and use that to manually send the "command"= ;, to see what happens. =A0It is able to execute correctly, seemingly infin= itely.

What I would like, is for the data stream to not just s= top at X amount of data transferred, I would like it to continue=A0infinite= ly similar to how it works when I manually send the command using Putty.
Please let me know if there is anything else I can provide for more hel= p!

Cheers!
Ben


--000e0cd32fd2ffdbe304a2662690-- --===============1144358550== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1144358550==-- From libssh2-devel-bounces@cool.haxx.se Wed May 4 00:14:53 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43MEe75030419; Wed, 4 May 2011 00:14:51 +0200 Received: from smtp02.local (mail.sss.co.nz [202.36.44.228]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p43MEZCu030366 for ; Wed, 4 May 2011 00:14:38 +0200 Message-ID: <4DC07E17.7080106@securitease.com> Date: Wed, 04 May 2011 10:13:43 +1200 From: Matt Lilley MIME-Version: 1.0 To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> <4DC06BDA.2040605@securitease.com> In-Reply-To: Content-Type: multipart/mixed; boundary="------------050805060307030407040301" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 00:14:53 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 00:14:38 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a multi-part message in MIME format. --------------050805060307030407040301 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed > I figured that, and I realize I'm arguing about a really odd edge-case > that probably won't ever happen to most users. I'm just striving to > make the API as clear as possible and offer as few surprises as > possible to the user of it. It also makes it easier to document and > describe. Yes, I can appreciate that. > > I'd like that if the user-set timeout is less than the keep-alive > timer has left, it should use the api_timeout and return an error if > it hits that timeout. > > If you update the patch, I'll offer to write the first draft of the > man pages for the two new functions! Deal ;-) I think this patch does that - but it's early in the morning still. Let me know what you think -- _____________________________________________ Matt Lilley Software Engineer SecuritEase Tel: +64 4 912-2100 Fax: +64 4 912-2101 E-mail: matt.lilley@securitease.com Web: http://www.securitease.com _____________________________________________ This e-mail has passed our content security scan. It is covered by the confidentiality clauses at http://www.securitease.com/content_and_confidentiality --------------050805060307030407040301 Content-Transfer-Encoding: 7bit Content-Type: text/plain; name="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" Content-Disposition: attachment; filename="0001-Fixes-bug-160-as-per-Daniel-s-suggestion.patch" From 684f6d16cfd46b535d29010e406f8fce457962b2 Mon Sep 17 00:00:00 2001 From: Matt Lilley Date: Tue, 3 May 2011 11:20:24 +1200 Subject: [PATCH] Fixes bug #160 as per Daniel's suggestion --- include/libssh2.h | 4 ++++ src/libssh2_priv.h | 3 +++ src/session.c | 52 +++++++++++++++++++++++++++++++++++++++++++++------- src/session.h | 44 +++++++++++++++++++++++--------------------- 4 files changed, 75 insertions(+), 28 deletions(-) diff --git a/include/libssh2.h b/include/libssh2.h index e207411..9462360 100644 --- a/include/libssh2.h +++ b/include/libssh2.h @@ -705,6 +705,10 @@ LIBSSH2_API int libssh2_session_get_blocking(LIBSSH2_SESSION* session); LIBSSH2_API void libssh2_channel_set_blocking(LIBSSH2_CHANNEL *channel, int blocking); +LIBSSH2_API void libssh2_session_set_timeout(LIBSSH2_SESSION* session, + long timeout); +LIBSSH2_API long libssh2_session_get_timeout(LIBSSH2_SESSION* session); + /* libssh2_channel_handle_extended_data is DEPRECATED, do not use! */ LIBSSH2_API void libssh2_channel_handle_extended_data(LIBSSH2_CHANNEL *channel, int ignore_mode); diff --git a/src/libssh2_priv.h b/src/libssh2_priv.h index 231401e..5a697c7 100644 --- a/src/libssh2_priv.h +++ b/src/libssh2_priv.h @@ -559,6 +559,9 @@ struct _LIBSSH2_SESSION /* this is set to TRUE if a blocking API behavior is requested */ int api_block_mode; + /* Timeout used when blocking API behavior is active */ + long api_timeout; + /* Server's public key */ const LIBSSH2_HOSTKEY_METHOD *hostkey; void *server_hostkey_abstract; diff --git a/src/session.c b/src/session.c index e4714e8..ef64cf6 100644 --- a/src/session.c +++ b/src/session.c @@ -481,6 +481,7 @@ libssh2_session_init_ex(LIBSSH2_ALLOC_FUNC((*my_alloc)), session->free = local_free; session->realloc = local_realloc; session->abstract = abstract; + session->api_timeout = 0; /* timeout-free API by default */ session->api_block_mode = 1; /* blocking API by default */ _libssh2_debug(session, LIBSSH2_TRACE_TRANS, "New session resource allocated"); @@ -542,11 +543,12 @@ libssh2_session_callback_set(LIBSSH2_SESSION * session, * Utility function that waits for action on the socket. Returns 0 when ready * to run again or error on timeout. */ -int _libssh2_wait_socket(LIBSSH2_SESSION *session) +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t start_time) { int rc; int seconds_to_next; int dir; + int has_timeout; /* since libssh2 often sets EAGAIN internally before this function is called, we can decrease some amount of confusion in user programs by @@ -591,9 +593,25 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) fd_set *writefd = NULL; fd_set *readfd = NULL; struct timeval tv; - - tv.tv_sec = seconds_to_next; - tv.tv_usec = 0; + + if (session->api_timeout > 0 && + (seconds_to_next == 0 || seconds_to_next > session->api_timeout)) { + time_t now = time (NULL); + long elapsed_ms = (long)(1000*difftime(start_time, now)); + if (elapsed_ms > session->api_timeout) { + session->err_code = LIBSSH2_ERROR_TIMEOUT; + return LIBSSH2_ERROR_TIMEOUT; + } + tv.tv_sec = (session->api_timeout - elapsed_ms) / 1000; + tv.tv_usec = ((session->api_timeout - elapsed_ms) % 1000) * 1000; + has_timeout = 1; + } else if (seconds_to_next > 0) { + tv.tv_sec = seconds_to_next; + tv.tv_usec = 0; + has_timeout = 1; + } else { + has_timeout = 0; + } if(dir & LIBSSH2_SESSION_BLOCK_INBOUND) { FD_ZERO(&rfd); @@ -607,10 +625,8 @@ int _libssh2_wait_socket(LIBSSH2_SESSION *session) writefd = &wfd; } - /* Note that this COULD be made to use a timeout that perhaps - could be customizable by the app or something... */ rc = select(session->socket_fd + 1, readfd, writefd, NULL, - seconds_to_next ? &tv : NULL); + has_timeout ? &tv : NULL); #endif } } @@ -1281,6 +1297,28 @@ libssh2_session_get_blocking(LIBSSH2_SESSION * session) return session->api_block_mode; } + +/* libssh2_session_set_timeout + * + * Set a session's timeout (in msec) for blocking mode, + * or 0 to disable timeouts. + */ +LIBSSH2_API void +libssh2_session_set_timeout(LIBSSH2_SESSION * session, long timeout) +{ + session->api_timeout = timeout; +} + +/* libssh2_session_get_timeout + * + * Returns a session's timeout, or 0 if disabled + */ +LIBSSH2_API long +libssh2_session_get_timeout(LIBSSH2_SESSION * session) +{ + return session->api_timeout; +} + /* * libssh2_poll_channel_read * diff --git a/src/session.h b/src/session.h index 4eac3a6..2866fcb 100644 --- a/src/session.h +++ b/src/session.h @@ -53,15 +53,16 @@ */ #define BLOCK_ADJUST(rc,sess,x) \ do { \ - rc = x; \ - /* the order of the check below is important to properly deal with the - case when the 'sess' is freed */ \ - if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + rc = x; \ + /* the order of the check below is important to properly deal with the + case when the 'sess' is freed */ \ + if((rc != LIBSSH2_ERROR_EAGAIN) || !sess->api_block_mode) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(!rc); \ + } while(0) /* * For functions that returns a pointer, we need to check if the API is @@ -69,21 +70,22 @@ * immediately. If the API is blocking and we get a NULL we check the errno * and *only* if that is EAGAIN we loop and wait for socket action. */ -#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ +#define BLOCK_ADJUST_ERRNO(ptr,sess,x) \ do { \ - int rc; \ - ptr = x; \ - if(!sess->api_block_mode || \ - (ptr != NULL) || \ - (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ - break; \ - rc = _libssh2_wait_socket(sess); \ - if(rc) \ - break; \ - } while(1) + time_t entry_time = time (NULL); \ + do { \ + int rc; \ + ptr = x; \ + if(!sess->api_block_mode || \ + (ptr != NULL) || \ + (libssh2_session_last_errno(sess) != LIBSSH2_ERROR_EAGAIN) ) \ + break; \ + rc = _libssh2_wait_socket(sess, entry_time); \ + } while(!rc); \ + } while(0) -int _libssh2_wait_socket(LIBSSH2_SESSION *session); +int _libssh2_wait_socket(LIBSSH2_SESSION *session, time_t entry_time); /* this is the lib-internal set blocking function */ int _libssh2_session_set_blocking(LIBSSH2_SESSION * session, int blocking); -- 1.7.2.3.msysgit.0 --------------050805060307030407040301 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --------------050805060307030407040301-- From libssh2-devel-bounces@cool.haxx.se Wed May 4 02:55:13 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p440sns5015988; Wed, 4 May 2011 02:55:09 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p440slSH015974 for ; Wed, 4 May 2011 02:54:47 +0200 Received: (qmail 26307 invoked by uid 501); 4 May 2011 00:54:54 -0000 Message-ID: <20110504005454.26306.qmail@stuge.se> Date: Wed, 4 May 2011 02:54:54 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: libssh2_channel_exec guidance Mail-Followup-To: libssh2-devel@cool.haxx.se References: MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 02:55:13 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 02:54:47 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se bayan wrote: > ssh_channel = libssh2_channel_open_session(ssh_session); > libssh2_channel_exec(m_ssh_channel, command); .. > Everything is working all fine and dandy (oh ya, dandy!), up until I > transfer X amount of (large?) data. The amount of data transferred is > exactly the same every time, at which point it completely halts all > transfer. .. > Please let me know if there is anything else I can provide for more help! For starters make X known. It would also be very useful if you could enable tracing and provide the output from that. (You may have to build a version of libssh2 that includes the debug messages.) //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 4 09:01:55 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4471PA2015812; Wed, 4 May 2011 09:01:49 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4471NRF015772 for ; Wed, 4 May 2011 09:01:23 +0200 Date: Wed, 4 May 2011 09:01:23 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: libssh2_channel_exec guidance In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 09:01:55 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 09:01:23 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 3 May 2011, bayan wrote: > I have a question that I have been unable to figure out, after searching > around and not coming up with anything, I was hoping to get guidance from > the hive-mind. It'll also be very useful to know which libssh2 version on what platform you're using! -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 4 11:05:35 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4495Jul020015; Wed, 4 May 2011 11:05:33 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4495IYC019996 for ; Wed, 4 May 2011 11:05:18 +0200 Received: (qmail 17296 invoked from network); 4 May 2011 09:05:18 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 4 May 2011 09:05:18 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Wed, 04 May 2011 09:05:18 -0000 X-URL: http://trac.libssh2.org/ Subject: [libssh2] #217: not multi-threaded safe X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/217 Message-ID: <048.be9e45dd6f70415e904487518592a6b7@libssh2.stuge.se> X-Trac-Ticket-ID: 217 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 11:05:35 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 11:05:18 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4495IYC019996 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #217: not multi-threaded safe ---------------------------+------------------------------------------------ Reporter: jianjingchen | Owner: Type: defect | Status: new Priority: normal | Milestone: 1.2.9 Component: API | Version: 1.2.8 Keywords: | Blocks: Blocked By: | ---------------------------+------------------------------------------------ I find function make_ctr_evp in file openssl.c is not is not thread safe, only one EVP_CIPHER is used for some threads concurrently. Function _libssh2_transport_read in file transport.c use function decrypt two times. for first time, p->payload should be NULL, then LIBSSH2_FREE(session, p->payload) will caught exception. -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 4 11:09:49 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4499lDf023479; Wed, 4 May 2011 11:09:49 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4499ktk023454 for ; Wed, 4 May 2011 11:09:46 +0200 Received: (qmail 18012 invoked from network); 4 May 2011 09:09:49 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 4 May 2011 09:09:49 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Wed, 04 May 2011 09:09:49 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #217: not multi-threaded safe X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/217#comment:1 Message-ID: <063.498c4a2427a7a525c604b9a2dc9af5dd@libssh2.stuge.se> References: <048.be9e45dd6f70415e904487518592a6b7@libssh2.stuge.se> X-Trac-Ticket-ID: 217 In-Reply-To: <048.be9e45dd6f70415e904487518592a6b7@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 11:09:49 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 11:09:46 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4499ktk023454 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #217: not multi-threaded safe ---------------------------+------------------------------------------------ Reporter: jianjingchen | Owner: Type: defect | Status: closed Priority: normal | Milestone: 1.2.9 Component: API | Version: 1.2.8 Resolution: invalid | Keywords: Blocks: | Blocked By: ---------------------------+------------------------------------------------ Changes (by bagder): * status: new => closed * resolution: => invalid Comment: OpenSSL requires the application to register two mutex callbacks when used multi-threaded, and if you do so then libssh2 should work fine. This is not a libssh2 bug. -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 4 18:04:42 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44G4BhP001168; Wed, 4 May 2011 18:04:37 +0200 Received: from mail-pz0-f54.google.com (mail-pz0-f54.google.com [209.85.210.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44G477s000860 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Wed, 4 May 2011 18:04:08 +0200 Received: by pzk27 with SMTP id 27so707725pzk.41 for ; Wed, 04 May 2011 09:04:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=AnsXs/grLk65y7qfRhPX5JSQn9vqW3p+HS5MyLhhd+k=; b=p4gbTAVix5PSMC0tLXJ2aJ4MuuL5FNhAzxNwwbaAu2ieeNUAu6da0Y0H1GEYN46NdF Tm2HMUd/6jtmzHBEiEt7JiwXtHELo+yG5BvKMw2/+UpODL+PQwyWpGOh5OQapLiROFWG /jt9suh1pTk3+PTTTkgfRMvidCm3czHeiAgr4= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=k41vbYNPi1cQktGgxARSpP6sdwwfFa5vK75QTjqMZ86ecXArJnpK6bz89DQSBv1k8C rAnQkpZ9YLD/ESEbSV3uycDlh/avV/YBmZWBgFoWXnrRShaaTwRnrcEcRMZe79RTY3XL o5o0+SrUV2/v6miLfmQdpxNAI63TnipMLBy0E= MIME-Version: 1.0 Received: by 10.142.62.42 with SMTP id k42mr685986wfa.230.1304525043301; Wed, 04 May 2011 09:04:03 -0700 (PDT) Received: by 10.142.212.16 with HTTP; Wed, 4 May 2011 09:04:03 -0700 (PDT) In-Reply-To: <20110504005454.26306.qmail@stuge.se> References: <20110504005454.26306.qmail@stuge.se> Date: Wed, 4 May 2011 12:04:03 -0400 Message-ID: Subject: Re: libssh2_channel_exec guidance From: bayan To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 18:04:42 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 18:04:09 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1162767767==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1162767767== Content-Type: multipart/alternative; boundary=001636e0adb8ce82e004a2756758 --001636e0adb8ce82e004a2756758 Content-Type: text/plain; charset=ISO-8859-1 Hi Peter, Thanks for the reply! I will go ahead and attempt the tracing compile, hopefully that will provide some more information that i can post when I get a chance. For now, I kept track of the number of characters transferred each time which totals to ~73000 characters (give or take a few, I suppose due to the way I am measuring it) Replacing my "command" libssh2 exec call, with simply putting "command" into putty, I am able to keep going well beyond this 73k character point where it dies through libssh2 everytime. As Daniel suggested, here is some more information: libssh2 version: #define LIBSSH2_VERSION "1.2.7" libssh2 compiled and running on Windows XP, connecting to a Unix machine. Cheers! Ben On Tue, May 3, 2011 at 8:54 PM, Peter Stuge wrote: > bayan wrote: > > ssh_channel = libssh2_channel_open_session(ssh_session); > > libssh2_channel_exec(m_ssh_channel, command); > .. > > Everything is working all fine and dandy (oh ya, dandy!), up until I > > transfer X amount of (large?) data. The amount of data transferred is > > exactly the same every time, at which point it completely halts all > > transfer. > .. > > Please let me know if there is anything else I can provide for more help! > > For starters make X known. > > It would also be very useful if you could enable tracing and provide > the output from that. (You may have to build a version of libssh2 > that includes the debug messages.) > > > //Peter > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > --001636e0adb8ce82e004a2756758 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Hi Peter,

Thanks for the reply!

I will go ahead and attempt = the tracing compile, hopefully that will provide some more information that= i can post when I get a chance.

For now, I kept track of the number= of characters transferred each time which totals to ~73000 characters (giv= e or take a few, I suppose due to the way I am measuring it)

Replacing my "command" libssh2 exec call, with simply putting= "command" into putty, I am able to keep going well beyond this 7= 3k character point where it dies through libssh2 everytime.

As Daniel suggested, here is some more information:

libs= sh2 version: #define LIBSSH2_VERSION "1.2.7"

=
libssh2 compiled and running on Windows XP, connecting to a Unix machi= ne.

Cheers!
Ben

On Tue, May 3, 2011 at 8:54 PM, Peter Stuge <peter@stuge.se> wrote:
bayan wrote:
> ssh_channel =3D libssh2_channel_open_session(ssh_session);
> libssh2_channel_exec(m_ssh_channel, command);
..
> Everything is working all fine and dandy (oh ya, dandy!), up until I > transfer X amount of (large?) data. =A0The amount of data transferred = is
> exactly the same every time, at which point it completely halts all > transfer.
..
> Please let me know if there is anything else I can provide for more he= lp!

For starters make X known.

It would also be very useful if you could enable tracing and provide
the output from that. (You may have to build a version of libssh2
that includes the debug messages.)


//Peter
_______________________________________________
libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/li= bssh2-devel

--001636e0adb8ce82e004a2756758-- --===============1162767767== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1162767767==-- From libssh2-devel-bounces@cool.haxx.se Wed May 4 18:37:03 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44GatEX030328; Wed, 4 May 2011 18:37:01 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44GaplX030270 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Wed, 4 May 2011 18:36:53 +0200 Received: by pvg11 with SMTP id 11so724684pvg.41 for ; Wed, 04 May 2011 09:36:51 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=04khK/JRWvCbu4QbFxGTwTfcinBFL8+4E4EN6Vq7g/U=; b=p/E+hPUGPsOmBt49ITxIJ+rgL+yDi3+h3COmP+xEfU2NG38qsy58s/Qt3xEfAqwsM/ TLzOqDbozZ5f4Wuk4AucKIuCGQXWrtLYz62j2rDVJoeKjoWsbjl/vOC9SiVdVRgHMwkP Bq4c+eZg9CnF8tc5XwnAeI2q8Ossof+ahQL28= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=m87TFsnfWH03KeXoJkWG0trcd44u2/9BCBB1EcL+eZjgrPwkhr3/SLLBT7AYPfStIx x3tfi1CjUs3pT8BU92BIWPd+8+suMq0LOa+EEMqcCPxjUlv7TveBHXgygVlxJVeiP2fp +RzjLk/l14n81RMCzFOljFqvlrA7HI/U5R4Rs= MIME-Version: 1.0 Received: by 10.142.52.8 with SMTP id z8mr646864wfz.401.1304527011408; Wed, 04 May 2011 09:36:51 -0700 (PDT) Received: by 10.142.212.16 with HTTP; Wed, 4 May 2011 09:36:51 -0700 (PDT) In-Reply-To: References: <20110504005454.26306.qmail@stuge.se> Date: Wed, 4 May 2011 12:36:51 -0400 Message-ID: Subject: Re: libssh2_channel_exec guidance From: bayan To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 18:37:03 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 18:36:53 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1246390308==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1246390308== Content-Type: multipart/alternative; boundary=000e0cd2e1101d6dfd04a275ddf8 --000e0cd2e1101d6dfd04a275ddf8 Content-Type: text/plain; charset=ISO-8859-1 Major typo on my part, the number of characters transferred needs another 0 appended to it: 730000 Its not 73k, its 730k. It transfers quite a decent amount before the disconnect. I also just attempted using a shell and write() instead of using exec, and it has the same behavior. I will compile a debug version to get some trace statements pumped out! On Wed, May 4, 2011 at 12:04 PM, bayan wrote: > Hi Peter, > > Thanks for the reply! > > I will go ahead and attempt the tracing compile, hopefully that will > provide some more information that i can post when I get a chance. > > For now, I kept track of the number of characters transferred each time > which totals to ~73000 characters (give or take a few, I suppose due to the > way I am measuring it) > > Replacing my "command" libssh2 exec call, with simply putting "command" > into putty, I am able to keep going well beyond this 73k character point > where it dies through libssh2 everytime. > > As Daniel suggested, here is some more information: > > libssh2 version: #define LIBSSH2_VERSION "1.2.7" > > libssh2 compiled and running on Windows XP, connecting to a Unix machine. > > Cheers! > Ben > > On Tue, May 3, 2011 at 8:54 PM, Peter Stuge wrote: > >> bayan wrote: >> > ssh_channel = libssh2_channel_open_session(ssh_session); >> > libssh2_channel_exec(m_ssh_channel, command); >> .. >> > Everything is working all fine and dandy (oh ya, dandy!), up until I >> > transfer X amount of (large?) data. The amount of data transferred is >> > exactly the same every time, at which point it completely halts all >> > transfer. >> .. >> > Please let me know if there is anything else I can provide for more >> help! >> >> For starters make X known. >> >> It would also be very useful if you could enable tracing and provide >> the output from that. (You may have to build a version of libssh2 >> that includes the debug messages.) >> >> >> //Peter >> _______________________________________________ >> libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel >> > > --000e0cd2e1101d6dfd04a275ddf8 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Major typo on my part, the number of characters=A0transferred=A0needs anoth= er 0 appended to it: 730000

Its not 73k, its 730k. =A0It transfers q= uite a decent amount before the disconnect.

I also just = attempted using a shell and write() instead of using exec, and it has the s= ame behavior.

I will compile a debug version to get some trace statements pumped out!=

On Wed, May 4, 2011 at 12:04 PM, bayan <= span dir=3D"ltr"><nayab9@gmail.com> wrote:
Hi Peter,

Thanks for the reply!
=
I will go ahead and attempt the tracing compile, hopefully that will pr= ovide some more information that i can post when I get a chance.

For now, I kept track of the number of characters transferred each time= which totals to ~73000 characters (give or take a few, I suppose due to th= e way I am measuring it)

Replacing my "command" libssh2 exec call, with simply putting= "command" into putty, I am able to keep going well beyond this 7= 3k character point where it dies through libssh2 everytime.

As Daniel suggested, here is some more information:

libs= sh2 version: #define LIBSSH2_VERSION "1.2.7"

=
libssh2 compiled and running on Windows XP, connecting to a Unix machi= ne.

Cheers!
Ben

On Tue, May 3, 2011 at 8:54 PM, Peter = Stuge <peter@stuge.se> wrote:
bayan wrote:
> ssh_channel =3D libssh2_channel_open_session(ssh_session);
> libssh2_channel_exec(m_ssh_channel, command);
..
> Everything is working all fine and dandy (oh ya, dandy!), up until I > transfer X amount of (large?) data. =A0The amount of data transferred = is
> exactly the same every time, at which point it completely halts all > transfer.
..
> Please let me know if there is anything else I can provide for more he= lp!

For starters make X known.

It would also be very useful if you could enable tracing and provide
the output from that. (You may have to build a version of libssh2
that includes the debug messages.)


//Peter
_______________________________________________
libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/li= bssh2-devel


--000e0cd2e1101d6dfd04a275ddf8-- --===============1246390308== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1246390308==-- From libssh2-devel-bounces@cool.haxx.se Wed May 4 19:06:06 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44H5mRn026889; Wed, 4 May 2011 19:06:05 +0200 Received: from foo.stuge.se (qmailr@foo.stuge.se [212.181.44.140]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with SMTP id p44H5koW026857 for ; Wed, 4 May 2011 19:05:46 +0200 Received: (qmail 26399 invoked by uid 501); 4 May 2011 17:05:51 -0000 Message-ID: <20110504170551.26398.qmail@stuge.se> Date: Wed, 4 May 2011 19:05:51 +0200 From: Peter Stuge To: libssh2-devel@cool.haxx.se Subject: Re: libssh2_channel_exec guidance Mail-Followup-To: libssh2-devel@cool.haxx.se References: <20110504005454.26306.qmail@stuge.se> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 19:06:06 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 19:05:46 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se bayan wrote: > > bayan wrote: > > > The amount of data transferred is exactly the same every time, > > > at which point it completely halts all transfer. .. > I kept track of the number of characters transferred each time > which totals to ~73000 characters (give or take a few, So which is it? Exactly the same every time, or give or take a few? Please try not to contradict yourself. > I suppose due to the way I am measuring it) This makes absolutely no sense at all. We are talking about discrete entities (bytes) and there can really only be one way to measure what you have sent. There are two metrics that would be interesting: 1. How many bytes of data you have passed to libssh2 on the channel. 2. How many bytes are succesfully received by the server side. Nothing but byte level detail is useful. > As Daniel suggested, here is some more information: > > libssh2 version: #define LIBSSH2_VERSION "1.2.7" Ok. This should be a sane version, but it would still be good for you to try out the latest code in git. > libssh2 compiled and running on Windows XP, connecting to a Unix machine. "Unix" doesn't say much. In particular it says nothing about the relevant software; the ssh daemon. OpenSSH-x.ypz with $distro patches is the level information you need to provide. It could also be very useful to provide debug output from the SSH server. //Peter _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Thu May 5 01:07:44 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44LS2OW020350; Wed, 4 May 2011 23:28:18 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p44LS1aP020293 for ; Wed, 4 May 2011 23:28:01 +0200 Date: Wed, 4 May 2011 23:28:00 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DC07E17.7080106@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> <4DC06BDA.2040605@securitease.com> <4DC07E17.7080106@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 04 May 2011 23:28:21 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 04 May 2011 23:28:01 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Wed, 4 May 2011, Matt Lilley wrote: > I think this patch does that - but it's early in the morning still. Let me > know what you think Thanks, I've now committed and pushed your patch with my little fix and two new man pages for the new functions. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Thu May 5 19:11:08 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p45HAaxL024199; Thu, 5 May 2011 19:11:04 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p45HAZRo024191 for ; Thu, 5 May 2011 19:10:35 +0200 Received: (qmail 6615 invoked from network); 5 May 2011 17:10:38 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 5 May 2011 17:10:38 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Thu, 05 May 2011 17:10:38 -0000 X-URL: http://trac.libssh2.org/ Subject: [libssh2] #218: libssh2_session_startup fails with certain ssh servers X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/218 Message-ID: <042.09f759c90c714b7de56dcc47f5be8529@libssh2.stuge.se> X-Trac-Ticket-ID: 218 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Thu, 05 May 2011 19:11:08 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Thu, 05 May 2011 19:10:35 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p45HAZRo024191 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #218: libssh2_session_startup fails with certain ssh servers ---------------------+------------------------------------------------------ Reporter: mstrsn | Owner: Type: defect | Status: new Priority: normal | Milestone: 1.2.9 Component: SFTP | Version: 1.2.8 Keywords: banner | Blocks: Blocked By: | ---------------------+------------------------------------------------------ Using libssh2_session_startup() with one particular ssh server always fails with LIBSSH2_ERROR_BANNER_RECV. Debugging the problem, I learned that my client program does not receive the complete banner from the server before blocking. This is a problem, because the loop in session_startup() that calls banner_receive() resets banner_TxRx_state to libssh2_NB_state_idle within the loop, causing the first portion of the banner to be lost when banner_receive() tries to get the rest of the banner. The fix that works for me is to update session.c like this: {{{ #!diff --- session.c.orig Tue Apr 26 13:41:46 2011 +++ session.c Tue Apr 26 14:59:06 2011 @@ -657,12 +657,11 @@ "Failed sending banner"); } session->startup_state = libssh2_NB_state_sent; + session->banner_TxRx_state = libssh2_NB_state_idle; } if (session->startup_state == libssh2_NB_state_sent) { do { - session->banner_TxRx_state = libssh2_NB_state_idle; - rc = banner_receive(session); if (rc) return _libssh2_error(session, rc, }}} I am using Solaris 10 x386 with gcc 3.4.6. The ssh server is running on a Ruggedcom switch running ROS 3.8.1. The banner returned by this server is "SSH-2.0-Mocana SSH". -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 10 08:48:07 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4A6le3t006708; Tue, 10 May 2011 08:48:03 +0200 Received: from mail-iw0-f182.google.com (mail-iw0-f182.google.com [209.85.214.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4A6lbHq006694 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 10 May 2011 08:47:38 +0200 Received: by iwn33 with SMTP id 33so6630829iwn.41 for ; Mon, 09 May 2011 23:47:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:date:message-id:subject:from:to :content-type; bh=QFuesEC1E7xILdql34aWoIWg2hjsFzcuKvlACGkwJ34=; b=PoefQoCURpr7QAvdRUVO3EO9K6azFcz8bHXzhhRHNM8CsHvDINPyo8wQ8m3x9tgZiL 8bXw1FpQXvMjN801Ypkn2CDyZMyPqdGyD5f+F9aFbLY+pHNyw0yenV05A5UEG/apMUE1 608GDzZZ1dm0POYCBg2/YDLQge7MTDM3IhjeM= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=m/YM/UoAV8hDT16hBSQeVt1aBjCDIB05jYLfdHLCcBUS3NTWkXrMqw76LeqgTwhm17 FOB07zRN1kwWchqHUDGFNdhDTGeijlzz77Jaz+GCSLpz0Jx4wyk5V7oLb+ID/sxQ3kp7 P/xe90keOq5ijiC9wHhbMbOjQ4/ACE+JxmcuM= MIME-Version: 1.0 Received: by 10.231.184.156 with SMTP id ck28mr5127140ibb.141.1305010052161; Mon, 09 May 2011 23:47:32 -0700 (PDT) Received: by 10.231.157.77 with HTTP; Mon, 9 May 2011 23:47:32 -0700 (PDT) Date: Tue, 10 May 2011 14:47:32 +0800 Message-ID: Subject: sftp_write_sliding.c bug From: zl liu To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 10 May 2011 08:48:07 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 10 May 2011 08:47:39 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1484620801==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1484620801== Content-Type: multipart/alternative; boundary=0050450174aa96607f04a2e65440 --0050450174aa96607f04a2e65440 Content-Type: text/plain; charset=GB2312 Content-Transfer-Encoding: quoted-printable Hi all=A3=BA sftp_write_sliding.c=A3=BA line 237 I modified like this: nread =3D fread(&mem[memuse], 1, sizeof(mem)-memuse, local); if (nread <=3D 0) { /* end of file */ if (memuse > 0){ /* the previous sending is not finished */ nread =3D 0; } else{ break; } } --0050450174aa96607f04a2e65440 Content-Type: text/html; charset=GB2312 Content-Transfer-Encoding: quoted-printable Hi all=A3=BA
sftp_write_sliding.c=A3=BA line 237
I = modified like this:
nread =3D fread(&mem[memuse], 1, siz= eof(mem)-memuse, local);
        if (nre= ad <=3D 0) {
            /*= end of file */
            if (memuse > 0){
                /* the = previous sending is not finished */
       &n= bsp;        nread =3D 0;
     =        }
         &n= bsp;  else{
             =    break;
            }
  = ;      }
--0050450174aa96607f04a2e65440-- --===============1484620801== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1484620801==-- From libssh2-devel-bounces@cool.haxx.se Tue May 10 10:18:34 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4A8II8m022216; Tue, 10 May 2011 10:18:33 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4A8IFBK022196 for ; Tue, 10 May 2011 10:18:15 +0200 Date: Tue, 10 May 2011 10:18:15 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: sftp_write_sliding.c bug In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="1129329158-1448261391-1305015495=:20293" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 10 May 2011 10:18:34 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 10 May 2011 10:18:15 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --1129329158-1448261391-1305015495=:20293 Content-Type: TEXT/PLAIN; charset=GB2312; format=flowed Content-Transfer-Encoding: 8BIT On Tue, 10 May 2011, zl liu wrote: > sftp_write_sliding.c£º line 237 > I modified like this: Thanks, committed! -- / daniel.haxx.se --1129329158-1448261391-1305015495=:20293 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --1129329158-1448261391-1305015495=:20293-- From libssh2-devel-bounces@cool.haxx.se Tue May 17 19:12:16 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHBLZI003335; Tue, 17 May 2011 19:12:12 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHBJRr003332 for ; Tue, 17 May 2011 19:11:19 +0200 Received: (qmail 7251 invoked from network); 17 May 2011 17:11:21 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 17 May 2011 17:11:21 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Tue, 17 May 2011 17:11:21 -0000 X-URL: http://trac.libssh2.org/ Subject: [libssh2] #219: libssh2_channel_forward_listen_ex freezes the program! X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/219 Message-ID: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Trac-Ticket-ID: 219 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 17 May 2011 19:12:16 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 17 May 2011 19:11:20 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4HHBJRr003332 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #219: libssh2_channel_forward_listen_ex freezes the program! ---------------------------+------------------------------------------------ Reporter: fingster | Owner: Type: defect | Status: new Priority: high | Milestone: 1.2.9 Component: API | Version: 1.2.8 Keywords: port forward | Blocks: Blocked By: | ---------------------------+------------------------------------------------ I use libssh2_channel_forward_listen_ex to implement the function of 'ssh -L ::'. however, libssh2_channel_forward_listen_ex stuck. The code is based on this example [http://stackoverflow.com/questions/1580750/example-code-of-libssh2-being- used-for-port-forwarding], like this: {{{ // initialize ssh connection int other_port = 0; LIBSSH2_LISTENER * fwd = libssh2_channel_forward_listen_ex(session, "127.0.0.1", 3128, &other_port, 1); printf("Remote forwding port: %d\n", other_port); channel = libssh2_channel_forward_accept(fwd); // pipe between localport and channel }}} -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 17 19:22:37 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHMUJr007583; Tue, 17 May 2011 19:22:36 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHMSSF007577 for ; Tue, 17 May 2011 19:22:28 +0200 Received: (qmail 8570 invoked from network); 17 May 2011 17:22:33 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 17 May 2011 17:22:33 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Tue, 17 May 2011 17:22:33 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #219: libssh2_channel_forward_listen_ex freezes the program! X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/219#comment:1 Message-ID: <059.51e5f4304f3cee13e6f097e18afe216f@libssh2.stuge.se> References: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Trac-Ticket-ID: 219 In-Reply-To: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 17 May 2011 19:22:37 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 17 May 2011 19:22:28 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4HHMSSF007577 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #219: libssh2_channel_forward_listen_ex freezes the program! -----------------------+---------------------------------------------------- Reporter: fingster | Owner: bagder Type: defect | Status: assigned Priority: high | Milestone: 1.2.9 Component: API | Version: 1.2.8 Resolution: | Keywords: port forward Blocks: | Blocked By: -----------------------+---------------------------------------------------- Changes (by bagder): * owner: => bagder * status: new => assigned Comment: libssh2_channel_forward_listen_ex() can indeed "get stuck" while waiting for data. Why is that wrong? If you want to avoid it getting stuck, use non-blocking mode. -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 17 19:36:11 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHa5wB013515; Tue, 17 May 2011 19:36:11 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHa4Fk013498 for ; Tue, 17 May 2011 19:36:04 +0200 Received: (qmail 10419 invoked from network); 17 May 2011 17:36:08 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 17 May 2011 17:36:08 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Tue, 17 May 2011 17:36:08 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #219: libssh2_channel_forward_listen_ex freezes the program! X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/219#comment:2 Message-ID: <059.743c4af88d4b72a9af3c415380634c3b@libssh2.stuge.se> References: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Trac-Ticket-ID: 219 In-Reply-To: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 17 May 2011 19:36:11 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 17 May 2011 19:36:04 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4HHa4Fk013498 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #219: libssh2_channel_forward_listen_ex freezes the program! -----------------------+---------------------------------------------------- Reporter: fingster | Owner: bagder Type: defect | Status: assigned Priority: high | Milestone: 1.2.9 Component: API | Version: 1.2.8 Resolution: | Keywords: port forward Blocks: | Blocked By: -----------------------+---------------------------------------------------- Comment (by fingster): When I put {{{ libssh2_session_set_blocking(session, 0); }}} before the cited example code, channel gets nothing but NULL. And printf shows that "Remote forwding port: 0". If I put {{{ libssh2_session_set_blocking(session, 0); }}} after the cited example code, the program freezes at libssh2_channel_forward_listen_ex and never returns. Nothing outputs. how to use libssh2_channel_forward_listen_ex? Any example? Thank you! -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 17 19:39:55 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHdrnO014364; Tue, 17 May 2011 19:39:54 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHdpTp014349 for ; Tue, 17 May 2011 19:39:51 +0200 Received: (qmail 11053 invoked from network); 17 May 2011 17:39:56 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 17 May 2011 17:39:56 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Tue, 17 May 2011 17:39:56 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #219: libssh2_channel_forward_listen_ex freezes the program! X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/219#comment:3 Message-ID: <059.c82c560a02ff9d7d1fcf3de663db1694@libssh2.stuge.se> References: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Trac-Ticket-ID: 219 In-Reply-To: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 17 May 2011 19:39:54 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 17 May 2011 19:39:51 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4HHdpTp014349 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #219: libssh2_channel_forward_listen_ex freezes the program! -----------------------+---------------------------------------------------- Reporter: fingster | Owner: bagder Type: defect | Status: assigned Priority: high | Milestone: 1.2.9 Component: API | Version: 1.2.8 Resolution: | Keywords: port forward Blocks: | Blocked By: -----------------------+---------------------------------------------------- Comment (by bagder): If you get NULL during non-blocking use you need to extract the error code to figure out if it is EAGAIN. I don't think we have any example for this, but we'd love to get one. Please consider using the mailing list instead for general help and advice as you don't seem to have identified any specific bug. -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 17 19:42:32 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHgSaA017694; Tue, 17 May 2011 19:42:31 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4HHgQH1017656 for ; Tue, 17 May 2011 19:42:26 +0200 Received: (qmail 11385 invoked from network); 17 May 2011 17:42:30 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 17 May 2011 17:42:30 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Tue, 17 May 2011 17:42:30 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #208: Facing an issue while SFTP copying on Solaris X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/208#comment:1 Message-ID: <058.dbaacdcce90a8ba66e966c17332861e6@libssh2.stuge.se> References: <043.39d3a69fd6a34156d1850ba5422ba148@libssh2.stuge.se> X-Trac-Ticket-ID: 208 In-Reply-To: <043.39d3a69fd6a34156d1850ba5422ba148@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 17 May 2011 19:42:32 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 17 May 2011 19:42:26 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4HHgQH1017656 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #208: Facing an issue while SFTP copying on Solaris ----------------------+----------------------------------------------------- Reporter: babi2ki | Owner: bagder Type: defect | Status: assigned Priority: high | Milestone: 1.2.9 Component: SFTP | Version: 1.2.7 Resolution: | Keywords: Blocks: | Blocked By: ----------------------+----------------------------------------------------- Changes (by bagder): * owner: => bagder * status: new => assigned * milestone: 1.2.7 => 1.2.9 Comment: Please build libssh2 to enable tracing and then run it with that enabled and show us the output so that we can better understand what's actually happening! -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 18 08:42:26 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4I6g2wO011903; Wed, 18 May 2011 08:42:23 +0200 Received: from earth.stuge.se (earth.stuge.se [212.116.89.126]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4I6g0q6011829 for ; Wed, 18 May 2011 08:42:00 +0200 Received: (qmail 3603 invoked from network); 18 May 2011 06:42:02 -0000 Received: from unknown (HELO earth.stuge.se) (127.0.0.1) by localhost with SMTP; 18 May 2011 06:42:02 -0000 MIME-Version: 1.0 From: "libssh2 Trac" X-Trac-Version: 0.12 Precedence: bulk Cc: libssh2-devel@cool.haxx.se Auto-Submitted: auto-generated X-Mailer: Trac 0.12, by Edgewall Software X-Trac-Project: libssh2 Date: Wed, 18 May 2011 06:42:02 -0000 X-URL: http://trac.libssh2.org/ Subject: Re: [libssh2] #219: libssh2_channel_forward_listen_ex freezes the program! X-Trac-Ticket-URL: http://trac.libssh2.org/ticket/219#comment:4 Message-ID: <059.bd06ddb0a4fefc15c66aaa49ff22b9cd@libssh2.stuge.se> References: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Trac-Ticket-ID: 219 In-Reply-To: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 18 May 2011 08:42:26 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 18 May 2011 08:42:00 +0200 (CEST) X-MIME-Autoconverted: from base64 to 8bit by giant.haxx.se id p4I6g0q6011829 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Reply-To: trac@libssh2.stuge.se, libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se #219: libssh2_channel_forward_listen_ex freezes the program! -----------------------+---------------------------------------------------- Reporter: fingster | Owner: bagder Type: defect | Status: closed Priority: high | Milestone: 1.2.9 Component: API | Version: 1.2.8 Resolution: invalid | Keywords: port forward Blocks: | Blocked By: -----------------------+---------------------------------------------------- Changes (by fingster): * status: assigned => closed * resolution: => invalid Comment: Replying to [comment:3 bagder]: > If you get NULL during non-blocking use you need to extract the error code to figure out if it is EAGAIN. > > I don't think we have any example for this, but we'd love to get one. > > Please consider using the mailing list instead for general help and advice as you don't seem to have identified any specific bug. After googling for a night and half day, the problem is finally resolved! libssh2_channel_forward_listen_ex is proper to implement 'ssh -R' not 'ssh -L'. For 'ssh -L', try "direct_tcpip" way. Thanks for bagder. -- Ticket URL: libssh2 C library for writing portable SSH2 clients _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 18 08:44:06 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4I6i40i014222; Wed, 18 May 2011 08:44:05 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4I6i2dl014170 for ; Wed, 18 May 2011 08:44:02 +0200 Date: Wed, 18 May 2011 08:44:02 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: implementing ssh -R vs ssh -L In-Reply-To: <059.bd06ddb0a4fefc15c66aaa49ff22b9cd@libssh2.stuge.se> Message-ID: References: <044.d06f6ca1fdc86dce6e752ef3a98d50dd@libssh2.stuge.se> <059.bd06ddb0a4fefc15c66aaa49ff22b9cd@libssh2.stuge.se> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 18 May 2011 08:44:06 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 18 May 2011 08:44:02 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Wed, 18 May 2011, libssh2 Trac wrote: [quote from ticket #219] > After googling for a night and half day, the problem is finally resolved! > libssh2_channel_forward_listen_ex is proper to implement 'ssh -R' not 'ssh > -L'. For 'ssh -L', try "direct_tcpip" way. Hi friends, Is there anything we can add to the documentation to make the above situation clearer or better described? -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 18 16:59:41 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4IExKwi023229; Wed, 18 May 2011 16:59:38 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4IExHfw023171 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Wed, 18 May 2011 16:59:19 +0200 Received: by pvg11 with SMTP id 11so934115pvg.41 for ; Wed, 18 May 2011 07:59:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:date:message-id:subject:from:to :content-type; bh=N21eJdj/LNQz7hCo/m2Fn13AtL69F1lLoybgiTBwORs=; b=ONZ8Nn4qPxxQvfL3AKdPFqZhJF/wig2BElpnPuWI4OmQDZP+1M+75oFVwI0+3dIqJM Yitpx1HsUkNSO+GxWN+fSChsCRP4SZ9Fn5Yjn+7EFWVbmsHbUE46GgQn70f5Utmz2ZpT i37PR3m/7QxTLxL56f9XicB1tk8WqVPOLrdwI= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=CcGUIIvinFzqQjPQh0X2gRpXcBfOf4C38B7KROGyemP+jawsxMleXZu80CulxgTxJK R6/DlzS7j7O9NKzk96JgKh+KtfkPQQeQU+nXl18Nm83ScGIODGboCpS79tRbIAp0Q9Wv tn0Dnxu5Ad7YEzF1bpktieJlJZ6G1p+1b5yX8= MIME-Version: 1.0 Received: by 10.68.1.35 with SMTP id 3mr3143001pbj.7.1305730756567; Wed, 18 May 2011 07:59:16 -0700 (PDT) Received: by 10.68.60.197 with HTTP; Wed, 18 May 2011 07:59:16 -0700 (PDT) Date: Wed, 18 May 2011 07:59:16 -0700 Message-ID: Subject: touch a file remotely From: Mark Roden To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 18 May 2011 16:59:41 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 18 May 2011 16:59:19 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Hi all, I'm looking for the ability to touch a file on a remote server where I'm logged in via scponly. I have sftp read/write access in the directory I need to write things. Right now, in order to touch a file on the remote system, I create it locally, and then send it as a zero-length file. There's nothing wrong with that, except that it just seems inefficient (and requires local cleanup steps, making sure that everything goes into the right place, etc). Is there a way to touch a file on a remote system through a libssh2 function call? Thanks, Mark _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 18 19:28:31 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4IHSC6Q002912; Wed, 18 May 2011 19:28:28 +0200 Received: from ponderosa.fch.wimsey.bc.ca (zz201111.cipherkey.net [209.53.201.111]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4IHS8UB002907 for ; Wed, 18 May 2011 19:28:09 +0200 Received: by coneharvesters.com via sendmail with stdio id for libssh2-devel@cool.haxx.se; Wed, 18 May 2011 10:28:09 -0700 (PDT) Date: Wed, 18 May 2011 10:28:08 -0700 From: Dan Fandrich To: libssh2-devel@cool.haxx.se Subject: Re: touch a file remotely Message-ID: <20110518172807.GA15502@coneharvesters.com> Mail-Followup-To: libssh2-devel@cool.haxx.se References: MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 18 May 2011 19:28:31 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 18 May 2011 19:28:10 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Wed, May 18, 2011 at 07:59:16AM -0700, Mark Roden wrote: > I'm looking for the ability to touch a file on a remote server where > I'm logged in via scponly. I have sftp read/write access in the > directory I need to write things. If by "touch" you mean to change the mtime of the file, then you can do that in the SFTP subsystem using the libssh2_sftp_stat_ex() call. >>> Dan _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Fri May 20 14:35:39 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KCYESp022204; Fri, 20 May 2011 14:35:35 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KCYDJk022197 for ; Fri, 20 May 2011 14:34:13 +0200 Date: Fri, 20 May 2011 14:34:13 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: trace always enabled? Message-ID: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Fri, 20 May 2011 14:35:39 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Fri, 20 May 2011 14:34:13 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Hi friends, I've been thinking about libssh2's ability to provide detailed traces when built with --enable-debug. It is a very useful feature and we ask for such traces almost at once for just about every bug report we get. How about we enable the trace feature for all builds unconditionally? It will take away the huge hurdle for people that has to rebuild the lib etc and it will better allow apps to also be able to ask libssh2 to provide tracing etc. There might be some very minor performance penalties involved but we can probably make sure they are very small. I still think the gain in convenience will compensate more than enough. Any opinions, thoughts or other feedback? -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Fri May 20 14:44:23 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KChK1R028644; Fri, 20 May 2011 14:44:22 +0200 Received: from mail-iw0-f182.google.com (mail-iw0-f182.google.com [209.85.214.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KChH6k028505 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Fri, 20 May 2011 14:43:18 +0200 Received: by iwn33 with SMTP id 33so3966385iwn.41 for ; Fri, 20 May 2011 05:43:14 -0700 (PDT) MIME-Version: 1.0 Received: by 10.42.138.199 with SMTP id d7mr4183128icu.197.1305895394462; Fri, 20 May 2011 05:43:14 -0700 (PDT) Received: by 10.42.219.129 with HTTP; Fri, 20 May 2011 05:43:14 -0700 (PDT) X-Originating-IP: [32.164.160.108] Received: by 10.42.219.129 with HTTP; Fri, 20 May 2011 05:43:14 -0700 (PDT) In-Reply-To: References: Date: Fri, 20 May 2011 07:43:14 -0500 Message-ID: Subject: Re: trace always enabled? From: Neil Gierman To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Fri, 20 May 2011 14:44:23 +0200 (CEST) X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Fri, 20 May 2011 14:43:19 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0685573196==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============0685573196== Content-Type: multipart/alternative; boundary=90e6ba1efc1219ca5104a3b4772e --90e6ba1efc1219ca5104a3b4772e Content-Type: text/plain; charset=ISO-8859-1 On May 20, 2011 7:38 AM, "Daniel Stenberg" wrote: > > Hi friends, > > I've been thinking about libssh2's ability to provide detailed traces when built with --enable-debug. It is a very useful feature and we ask for such traces almost at once for just about every bug report we get. > > How about we enable the trace feature for all builds unconditionally? It will take away the huge hurdle for people that has to rebuild the lib etc and it will better allow apps to also be able to ask libssh2 to provide tracing etc. > > There might be some very minor performance penalties involved but we can probably make sure they are very small. I still think the gain in convenience will compensate more than enough. > > Any opinions, thoughts or other feedback? > > -- > > / daniel.haxx.se > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel I personally would like to keep it just for debug builds. I would really rather not have any extra debugging information compiled in the release builds even if it is only sitting dormant. --90e6ba1efc1219ca5104a3b4772e Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable


On May 20, 2011 7:38 AM, "Daniel Stenberg" <daniel@haxx.se> wrote:
>
> Hi friends,
>
> I've been thinking about libssh2's ability to provide detailed= traces when built with --enable-debug. It is a very useful feature and we = ask for such traces almost at once for just about every bug report we get.<= br> >
> How about we enable the trace feature for all builds unconditionally? = It will take away the huge hurdle for people that has to rebuild the lib et= c and it will better allow apps to also be able to ask libssh2 to provide t= racing etc.
>
> There might be some very minor performance penalties involved but we c= an probably make sure they are very small. I still think the gain in conven= ience will compensate more than enough.
>
> Any opinions, thoughts or other feedback?
>
> --
>
> =A0/ daniel.haxx.se
> _______________________________________________
> libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel

I personally would like to keep it just for debug builds. I would really= rather not have any extra debugging information compiled in the release bu= ilds even if it is only sitting dormant.

--90e6ba1efc1219ca5104a3b4772e-- --===============0685573196== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============0685573196==-- From libssh2-devel-bounces@cool.haxx.se Fri May 20 15:34:40 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KDYUM0021143; Fri, 20 May 2011 15:34:39 +0200 Received: from mail-px0-f177.google.com (mail-px0-f177.google.com [209.85.212.177]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KDYQA1021129 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Fri, 20 May 2011 15:34:28 +0200 Received: by pxi10 with SMTP id 10so2661690pxi.8 for ; Fri, 20 May 2011 06:34:23 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type:content-transfer-encoding; bh=rXaLCHBrVCpq+oqaZc4nclVk+SMKPBXCJosdByY5L94=; b=VZPhOq/kMzQF3FE8T8of1j213UIQZ388XWIcrUq1iPrP2CiiwzfuBG7MVdOleQm5xl yVTIPG1wTSX62qnimqPaI3mmrnxTbGIPbgz1vRnuYPXRWdXT+qSx8gq9KstOL0kDY6Io dLqktuuKUWv5ke5Qka6Os6Ty5qwGb02PY4RzY= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=vsB5KfUV3ndaiHnBynHn3V0HDeaXt5TPLXg2YLLq3vnY9Xjt9ryQ9YeWOCNvBrqDxE So3tFjvDlGmgj4KCHunI3Q5mFecqpOhWSjNhvfw4ftXRLRrZyIhWocg6eQpbKwGY3ima MP4/hcuKr9pCtibp9QGCwpbTQSHz//oz8j02k= MIME-Version: 1.0 Received: by 10.68.1.169 with SMTP id 9mr7043114pbn.88.1305898463614; Fri, 20 May 2011 06:34:23 -0700 (PDT) Received: by 10.68.52.72 with HTTP; Fri, 20 May 2011 06:34:23 -0700 (PDT) In-Reply-To: References: Date: Fri, 20 May 2011 06:34:23 -0700 Message-ID: Subject: Re: trace always enabled? From: Mark Roden To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Fri, 20 May 2011 15:34:40 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Fri, 20 May 2011 15:34:28 +0200 (CEST) X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p4KDYQA1021129 X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="iso-8859-1" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p4KDYUM0021143 I agree with Neil. You could also support a release-with-debug build, something that's optimized but still supports debugging macros. That way, if someone encounters a bug due to optimization, they at least will have traces available to them. On Fri, May 20, 2011 at 5:43 AM, Neil Gierman wrote: > > On May 20, 2011 7:38 AM, "Daniel Stenberg" wrote: >> >> Hi friends, >> >> I've been thinking about libssh2's ability to provide detailed traces when >> built with --enable-debug. It is a very useful feature and we ask for such >> traces almost at once for just about every bug report we get. >> >> How about we enable the trace feature for all builds unconditionally? It >> will take away the huge hurdle for people that has to rebuild the lib etc >> and it will better allow apps to also be able to ask libssh2 to provide >> tracing etc. >> >> There might be some very minor performance penalties involved but we can >> probably make sure they are very small. I still think the gain in >> convenience will compensate more than enough. >> >> Any opinions, thoughts or other feedback? >> >> -- >> >>  / daniel.haxx.se >> _______________________________________________ >> libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > > I personally would like to keep it just for debug builds. I would really > rather not have any extra debugging information compiled in the release > builds even if it is only sitting dormant. > > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > > _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Fri May 20 17:27:39 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KFQJmW014489; Fri, 20 May 2011 17:27:38 +0200 Received: from mx2.belnet.be (mx2.belnet.be [193.190.130.39]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4KFQHgF014451 for ; Fri, 20 May 2011 17:26:17 +0200 X-IronPort-AV: E=Sophos;i="4.65,242,1304287200"; d="scan'208";a="1217087" Received: from firewall.belnet.be (HELO office.belnet.be) ([193.190.130.1]) by mx2-out.belnet.be with ESMTP; 20 May 2011 17:26:16 +0200 Received: from localhost (localhost [127.0.0.1]) by office.belnet.be (Postfix) with ESMTP id C01BD4027F for ; Fri, 20 May 2011 17:26:16 +0200 (CEST) Received: from aris-laptop.local (unknown [10.4.99.174]) by office.belnet.be (Postfix) with ESMTP id B2EAF40279 for ; Fri, 20 May 2011 17:26:16 +0200 (CEST) Message-ID: <4DD68818.1090707@0xbadc0de.be> Date: Fri, 20 May 2011 17:26:16 +0200 From: Aris Adamantiadis User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; fr; rv:1.9.2.17) Gecko/20110414 Lightning/1.0b2 Thunderbird/3.1.10 MIME-Version: 1.0 To: libssh2-devel@cool.haxx.se Subject: Re: trace always enabled? References: In-Reply-To: X-Enigmail-Version: 1.1.1 X-Virus-Scanned: Belnet virus scanner X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Fri, 20 May 2011 17:27:39 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Fri, 20 May 2011 17:26:18 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="iso-8859-1" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by giant.haxx.se id p4KFQJmW014489 Le 20/05/11 15:34, Mark Roden a écrit : > I agree with Neil. > > You could also support a release-with-debug build, something that's > optimized but still supports debugging macros. That way, if someone > encounters a bug due to optimization, they at least will have traces > available to them. Hi, That's what we do with libssh, it's compiled by default, give us great debugging info when there's a problem (sometimes fixes it immediatly), and nobody ever complained about it taking too much CPU or space in the binary. Sometimes asking a user to recompile his distro's package is not realistic. Kr, Aris _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Sat May 21 12:58:45 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4LAw7GP030304; Sat, 21 May 2011 12:58:40 +0200 Received: from mail-ey0-f182.google.com (mail-ey0-f182.google.com [209.85.215.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4LAw4qc030219 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Sat, 21 May 2011 12:58:05 +0200 Received: by eyg7 with SMTP id 7so1959884eyg.41 for ; Sat, 21 May 2011 03:58:02 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=peut.org; s=google; h=domainkey-signature:message-id:date:from:user-agent:mime-version:to :subject:references:in-reply-to:content-type; bh=p3N204o0x+nlWGnmZw1Ii0NkcVZMSsRK+9/GfshUMEY=; b=KnAgf7iEm9u0yvC4RqNXogFSuLfXJUgZ2c1x4encgMofTcaSoGHqjLWwNEaWBXPaEp tFIyJYZy/706fWnKKvG7IFqG0sBVDrbPLZzNk4Qvn1rGHjFCNkEVO41K8ui+8tMyLYQ9 nkqiLxWCQHw5z3XookNQmCttNAO49Z2jq9ZSQ= DomainKey-Signature: a=rsa-sha1; c=nofws; d=peut.org; s=google; h=message-id:date:from:user-agent:mime-version:to:subject:references :in-reply-to:content-type; b=WbgYsuQL3Su59zJMz/Y7ds0w6Z0LuVBI1AIuNfsCRZntr+hlDMHcRUtAa8uWVAuya+ gqcwaGjMkywEDfFMGJIcNAOj4uD5UHoEPtcSYGLSr4ifv2kfOSEFvwsS2GcKUbM4LLyw yFTLkQYDSUQffhUQv7PNpDI8ODpB5z44d7fQQ= Received: by 10.213.9.145 with SMTP id l17mr474453ebl.10.1305975482569; Sat, 21 May 2011 03:58:02 -0700 (PDT) Received: from [192.168.0.35] (96-160.ftth.onsbrabantnet.nl [88.159.160.96]) by mx.google.com with ESMTPS id y3sm3198023eeh.19.2011.05.21.03.58.00 (version=SSLv3 cipher=OTHER); Sat, 21 May 2011 03:58:01 -0700 (PDT) Message-ID: <4DD79AB0.6030109@peut.org> Date: Sat, 21 May 2011 12:57:52 +0200 From: Jose Baars User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; nl; rv:1.9.2.17) Gecko/20110414 Thunderbird/3.1.10 MIME-Version: 1.0 To: libssh2-devel@cool.haxx.se Subject: Re: trace always enabled? References: In-Reply-To: X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Sat, 21 May 2011 12:58:45 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Sat, 21 May 2011 12:58:05 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1296556000==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This is a cryptographically signed message in MIME format. --===============1296556000== Content-Type: multipart/signed; protocol="application/pkcs7-signature"; micalg=sha1; boundary="------------ms080601050104070807090104" This is a cryptographically signed message in MIME format. --------------ms080601050104070807090104 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable In the VMS build stuff it's already enabled. With all the very different = implementations of SSH on very different systems around, you really can't do without a t= racing facility. My guess is the CPU penalty is below 0.1 percent*. Jos=E9 Baars* --------------ms080601050104070807090104 Content-Type: application/pkcs7-signature; name="smime.p7s" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="smime.p7s" Content-Description: S/MIME Cryptographic Signature MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAQAAoIIRtDCC BV0wggRFoAMCAQICEBUPxTeUYVibM15V7SdI4YcwDQYJKoZIhvcNAQEFBQAwgd0xCzAJBgNV BAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjEfMB0GA1UECxMWVmVyaVNpZ24gVHJ1 c3QgTmV0d29yazE7MDkGA1UECxMyVGVybXMgb2YgdXNlIGF0IGh0dHBzOi8vd3d3LnZlcmlz aWduLmNvbS9ycGEgKGMpMDkxHjAcBgNVBAsTFVBlcnNvbmEgTm90IFZhbGlkYXRlZDE3MDUG A1UEAxMuVmVyaVNpZ24gQ2xhc3MgMSBJbmRpdmlkdWFsIFN1YnNjcmliZXIgQ0EgLSBHMzAe Fw0xMDExMTIwMDAwMDBaFw0xMTExMTIyMzU5NTlaMIIBCjEXMBUGA1UEChMOVmVyaVNpZ24s IEluYy4xHzAdBgNVBAsTFlZlcmlTaWduIFRydXN0IE5ldHdvcmsxRjBEBgNVBAsTPXd3dy52 ZXJpc2lnbi5jb20vcmVwb3NpdG9yeS9SUEEgSW5jb3JwLiBieSBSZWYuLExJQUIuTFREKGMp OTgxHjAcBgNVBAsTFVBlcnNvbmEgTm90IFZhbGlkYXRlZDEzMDEGA1UECxMqRGlnaXRhbCBJ RCBDbGFzcyAxIC0gTmV0c2NhcGUgRnVsbCBTZXJ2aWNlMRMwEQYDVQQDFApKb3NlIEJhYXJz MRwwGgYJKoZIhvcNAQkBFg1wZXV0QHBldXQub3JnMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8A MIIBCgKCAQEAr/QJrg4jLwhCrNdHOzbC5HFzjE/FoklGXizIWpmW/bQOf6UQCVmJIPK7dsaJ Dz2zytkbrrre5s/qfPcNkSScKSOguxyBmnQ/tW6Y0OnlYzv0ZrYyC0J8zzntDfZ5mLOzBTCY rGS0Pqy8n7vRScDT927/62dEfgHiA3ceWbc+NCIJYdgupDjByypHgmi2H6JAiv6N16U2LRfM 2ZNBu4RYgWHQp31Y9cX/a/CHEv1rG/5MPVDQP6t45saHKJN/gcH4oXeCfHINZ7iUJhIWpk9i pGEjwPLpgRAsGJFRHrZTBgOjrwrdzfrDgnKWkB2RY+IAYQ7z+8+KJGHwfuWvsH2EbQIDAQAB o4HoMIHlMAkGA1UdEwQCMAAwRAYDVR0gBD0wOzA5BgtghkgBhvhFAQcXATAqMCgGCCsGAQUF BwIBFhxodHRwczovL3d3dy52ZXJpc2lnbi5jb20vcnBhMAsGA1UdDwQEAwIFoDAdBgNVHSUE FjAUBggrBgEFBQcDBAYIKwYBBQUHAwIwFAYKYIZIAYb4RQEGBwQGFgROb25lMFAGA1UdHwRJ MEcwRaBDoEGGP2h0dHA6Ly9pbmRjMWRpZ2l0YWxpZC1nMy1jcmwudmVyaXNpZ24uY29tL0lu ZEMxRGlnaXRhbElELUczLmNybDANBgkqhkiG9w0BAQUFAAOCAQEA5nI96tQkbCelTz4bLtnc 6uCIfbmE0slYOMLfmjwgXm0qJAMYCiBlhdBkBeROwP9JkqIRHMOnPAGB/i9nz/6KeIJQEBKT ClJMeU9NZ4VvIsmaTnEmL86fuDQygCYJXf7+9PfIbK2BDQ9AsJtvLPfQXfMkATYEAP/dozXX ZLiZ3p5sex3DzevIX8/TNjV4Vdt7S8rYXfLz+yZh5bJVliScankYwcScVh/0kn0DLdq47DxO ZANPISNkOF0yvfT/cTFMnLykCwtZiGGfRGl5KAl6DSN0vn49B+M0QuvGAN/5itaGx4YCLMl1 0iganL8H2SHFW4a+BeKZ7t+1JG9l/YUGdTCCBV0wggRFoAMCAQICEBUPxTeUYVibM15V7SdI 4YcwDQYJKoZIhvcNAQEFBQAwgd0xCzAJBgNVBAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwg SW5jLjEfMB0GA1UECxMWVmVyaVNpZ24gVHJ1c3QgTmV0d29yazE7MDkGA1UECxMyVGVybXMg b2YgdXNlIGF0IGh0dHBzOi8vd3d3LnZlcmlzaWduLmNvbS9ycGEgKGMpMDkxHjAcBgNVBAsT FVBlcnNvbmEgTm90IFZhbGlkYXRlZDE3MDUGA1UEAxMuVmVyaVNpZ24gQ2xhc3MgMSBJbmRp dmlkdWFsIFN1YnNjcmliZXIgQ0EgLSBHMzAeFw0xMDExMTIwMDAwMDBaFw0xMTExMTIyMzU5 NTlaMIIBCjEXMBUGA1UEChMOVmVyaVNpZ24sIEluYy4xHzAdBgNVBAsTFlZlcmlTaWduIFRy dXN0IE5ldHdvcmsxRjBEBgNVBAsTPXd3dy52ZXJpc2lnbi5jb20vcmVwb3NpdG9yeS9SUEEg SW5jb3JwLiBieSBSZWYuLExJQUIuTFREKGMpOTgxHjAcBgNVBAsTFVBlcnNvbmEgTm90IFZh bGlkYXRlZDEzMDEGA1UECxMqRGlnaXRhbCBJRCBDbGFzcyAxIC0gTmV0c2NhcGUgRnVsbCBT ZXJ2aWNlMRMwEQYDVQQDFApKb3NlIEJhYXJzMRwwGgYJKoZIhvcNAQkBFg1wZXV0QHBldXQu b3JnMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAr/QJrg4jLwhCrNdHOzbC5HFz jE/FoklGXizIWpmW/bQOf6UQCVmJIPK7dsaJDz2zytkbrrre5s/qfPcNkSScKSOguxyBmnQ/ tW6Y0OnlYzv0ZrYyC0J8zzntDfZ5mLOzBTCYrGS0Pqy8n7vRScDT927/62dEfgHiA3ceWbc+ NCIJYdgupDjByypHgmi2H6JAiv6N16U2LRfM2ZNBu4RYgWHQp31Y9cX/a/CHEv1rG/5MPVDQ P6t45saHKJN/gcH4oXeCfHINZ7iUJhIWpk9ipGEjwPLpgRAsGJFRHrZTBgOjrwrdzfrDgnKW kB2RY+IAYQ7z+8+KJGHwfuWvsH2EbQIDAQABo4HoMIHlMAkGA1UdEwQCMAAwRAYDVR0gBD0w OzA5BgtghkgBhvhFAQcXATAqMCgGCCsGAQUFBwIBFhxodHRwczovL3d3dy52ZXJpc2lnbi5j b20vcnBhMAsGA1UdDwQEAwIFoDAdBgNVHSUEFjAUBggrBgEFBQcDBAYIKwYBBQUHAwIwFAYK YIZIAYb4RQEGBwQGFgROb25lMFAGA1UdHwRJMEcwRaBDoEGGP2h0dHA6Ly9pbmRjMWRpZ2l0 YWxpZC1nMy1jcmwudmVyaXNpZ24uY29tL0luZEMxRGlnaXRhbElELUczLmNybDANBgkqhkiG 9w0BAQUFAAOCAQEA5nI96tQkbCelTz4bLtnc6uCIfbmE0slYOMLfmjwgXm0qJAMYCiBlhdBk BeROwP9JkqIRHMOnPAGB/i9nz/6KeIJQEBKTClJMeU9NZ4VvIsmaTnEmL86fuDQygCYJXf7+ 9PfIbK2BDQ9AsJtvLPfQXfMkATYEAP/dozXXZLiZ3p5sex3DzevIX8/TNjV4Vdt7S8rYXfLz +yZh5bJVliScankYwcScVh/0kn0DLdq47DxOZANPISNkOF0yvfT/cTFMnLykCwtZiGGfRGl5 KAl6DSN0vn49B+M0QuvGAN/5itaGx4YCLMl10iganL8H2SHFW4a+BeKZ7t+1JG9l/YUGdTCC Bu4wggXWoAMCAQICEHEVZgVK5JEhTem8RPms09wwDQYJKoZIhvcNAQEFBQAwgcoxCzAJBgNV BAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjEfMB0GA1UECxMWVmVyaVNpZ24gVHJ1 c3QgTmV0d29yazE6MDgGA1UECxMxKGMpIDE5OTkgVmVyaVNpZ24sIEluYy4gLSBGb3IgYXV0 aG9yaXplZCB1c2Ugb25seTFFMEMGA1UEAxM8VmVyaVNpZ24gQ2xhc3MgMSBQdWJsaWMgUHJp bWFyeSBDZXJ0aWZpY2F0aW9uIEF1dGhvcml0eSAtIEczMB4XDTA5MDUwMTAwMDAwMFoXDTE5 MDQzMDIzNTk1OVowgd0xCzAJBgNVBAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjEf MB0GA1UECxMWVmVyaVNpZ24gVHJ1c3QgTmV0d29yazE7MDkGA1UECxMyVGVybXMgb2YgdXNl IGF0IGh0dHBzOi8vd3d3LnZlcmlzaWduLmNvbS9ycGEgKGMpMDkxHjAcBgNVBAsTFVBlcnNv bmEgTm90IFZhbGlkYXRlZDE3MDUGA1UEAxMuVmVyaVNpZ24gQ2xhc3MgMSBJbmRpdmlkdWFs IFN1YnNjcmliZXIgQ0EgLSBHMzCCASIwDQYJKoZIhvcNAQEBBQADggEPADCCAQoCggEBAO3E R98qKB18Bmu71yEyyWwTj+mxjUFONPfaC+Nq+mWIIAsRE+mb4ElOi2/VAdBfDUeRilpMdD4/ xpEJu0w0no1uoYJRYvdpdliWB6+eFBgHT1q9n9IxslQZc0ZqGUIR7BJzIY313DDN5dlWCjHF Nm0pFJe9LdqJRxmI2EsEPeu2PGcedAATDdCG2pNn+DMDrho8a2l49sAsjuGDP3f5mf/+n1Ja wrSHCthsqUfBVCllQz5KwJYfwa33d69ssQRevsG2lC2XkC0n0rse6YNqhPbEsq4jBmUmpSdY KwcitG+mYkgad/LVUCeaKdOW+yj1uiR2YuOMWev7btVCxL5Bx/UCAwEAAaOCArkwggK1MDQG CCsGAQUFBwEBBCgwJjAkBggrBgEFBQcwAYYYaHR0cDovL29jc3AudmVyaXNpZ24uY29tMBIG A1UdEwEB/wQIMAYBAf8CAQAwcAYDVR0gBGkwZzBlBgtghkgBhvhFAQcXATBWMCgGCCsGAQUF BwIBFhxodHRwczovL3d3dy52ZXJpc2lnbi5jb20vY3BzMCoGCCsGAQUFBwICMB4aHGh0dHBz Oi8vd3d3LnZlcmlzaWduLmNvbS9ycGEwNAYDVR0fBC0wKzApoCegJYYjaHR0cDovL2NybC52 ZXJpc2lnbi5jb20vcGNhMS1nMy5jcmwwDgYDVR0PAQH/BAQDAgEGMG4GCCsGAQUFBwEMBGIw YKFeoFwwWjBYMFYWCWltYWdlL2dpZjAhMB8wBwYFKw4DAhoEFEtruSiWBgy70FI4mymsSweL IQUYMCYWJGh0dHA6Ly9sb2dvLnZlcmlzaWduLmNvbS92c2xvZ28xLmdpZjAuBgNVHREEJzAl pCMwITEfMB0GA1UEAxMWUHJpdmF0ZUxhYmVsNC0yMDQ4LTExODAdBgNVHQ4EFgQUeUdhCEH9 OASiS+e1zPVD9kkrEfgwgfEGA1UdIwSB6TCB5qGB0KSBzTCByjELMAkGA1UEBhMCVVMxFzAV BgNVBAoTDlZlcmlTaWduLCBJbmMuMR8wHQYDVQQLExZWZXJpU2lnbiBUcnVzdCBOZXR3b3Jr MTowOAYDVQQLEzEoYykgMTk5OSBWZXJpU2lnbiwgSW5jLiAtIEZvciBhdXRob3JpemVkIHVz ZSBvbmx5MUUwQwYDVQQDEzxWZXJpU2lnbiBDbGFzcyAxIFB1YmxpYyBQcmltYXJ5IENlcnRp ZmljYXRpb24gQXV0aG9yaXR5IC0gRzOCEQCLW3VWhFSFCwDPrzhIzrGkMA0GCSqGSIb3DQEB BQUAA4IBAQA5Tc9BmYG1qQW1UjjpOYSJbOQ0qFrn2GwJTCQaulmkhztzIfGTgc+/aGNaZ/41 hSuhw12jSsI6Gd0w1sxN7/HSgZfKVFpDvzeLeo4ZjQ9DqIzyr2CzFYqzlZw84J6zJ5ikNXIX 5fwqXYfTig3C0UUq+MD0rCqTOtWuEnAI6/s74nfs6CtkNXbNutrg0csU1nFYm77VPn222egk xSRmTF2RH3azFz5/DcYhiS+zN7ih/1yybUneZVJC+w6I0u1KHb9L4/jMcvpIDmWOScjW+JmY O7eUPjFxBof6bFlTLtffK+1fYwCsFe0DuFUWjMZoA+ciqHMLsbyg2lJY3QoOf8GCMYIE7DCC BOgCAQEwgfIwgd0xCzAJBgNVBAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjEfMB0G A1UECxMWVmVyaVNpZ24gVHJ1c3QgTmV0d29yazE7MDkGA1UECxMyVGVybXMgb2YgdXNlIGF0 IGh0dHBzOi8vd3d3LnZlcmlzaWduLmNvbS9ycGEgKGMpMDkxHjAcBgNVBAsTFVBlcnNvbmEg Tm90IFZhbGlkYXRlZDE3MDUGA1UEAxMuVmVyaVNpZ24gQ2xhc3MgMSBJbmRpdmlkdWFsIFN1 YnNjcmliZXIgQ0EgLSBHMwIQFQ/FN5RhWJszXlXtJ0jhhzAJBgUrDgMCGgUAoIICzjAYBgkq hkiG9w0BCQMxCwYJKoZIhvcNAQcBMBwGCSqGSIb3DQEJBTEPFw0xMTA1MjExMDU3NTJaMCMG CSqGSIb3DQEJBDEWBBT3zACx10X5pOdovbb0h12P1LX3WTBfBgkqhkiG9w0BCQ8xUjBQMAsG CWCGSAFlAwQBAjAKBggqhkiG9w0DBzAOBggqhkiG9w0DAgICAIAwDQYIKoZIhvcNAwICAUAw BwYFKw4DAgcwDQYIKoZIhvcNAwICASgwggEDBgkrBgEEAYI3EAQxgfUwgfIwgd0xCzAJBgNV BAYTAlVTMRcwFQYDVQQKEw5WZXJpU2lnbiwgSW5jLjEfMB0GA1UECxMWVmVyaVNpZ24gVHJ1 c3QgTmV0d29yazE7MDkGA1UECxMyVGVybXMgb2YgdXNlIGF0IGh0dHBzOi8vd3d3LnZlcmlz aWduLmNvbS9ycGEgKGMpMDkxHjAcBgNVBAsTFVBlcnNvbmEgTm90IFZhbGlkYXRlZDE3MDUG A1UEAxMuVmVyaVNpZ24gQ2xhc3MgMSBJbmRpdmlkdWFsIFN1YnNjcmliZXIgQ0EgLSBHMwIQ FQ/FN5RhWJszXlXtJ0jhhzCCAQUGCyqGSIb3DQEJEAILMYH1oIHyMIHdMQswCQYDVQQGEwJV UzEXMBUGA1UEChMOVmVyaVNpZ24sIEluYy4xHzAdBgNVBAsTFlZlcmlTaWduIFRydXN0IE5l dHdvcmsxOzA5BgNVBAsTMlRlcm1zIG9mIHVzZSBhdCBodHRwczovL3d3dy52ZXJpc2lnbi5j b20vcnBhIChjKTA5MR4wHAYDVQQLExVQZXJzb25hIE5vdCBWYWxpZGF0ZWQxNzA1BgNVBAMT LlZlcmlTaWduIENsYXNzIDEgSW5kaXZpZHVhbCBTdWJzY3JpYmVyIENBIC0gRzMCEBUPxTeU YVibM15V7SdI4YcwDQYJKoZIhvcNAQEBBQAEggEAS0Co1QedwFP8fW0fG9yl3LEuCZfbiaX9 xWOpspNdQu50qPs9H+BOqzFQadxzAWZRl40lKztnn2A9veKMH6IGrjzK3xHm3Zdr9Tfo9gbc 5hHnMY/Fc0jtoQW7IjOrLjyIwZzz8P+emTkR01Tpw/lIFCmp4jod6bN1Ifq8x+b4gvNtYWHc +Ob+Di89rW72BMho0X8qltik7jobQFiwgsDoWkshjtcgNGGI0zOuBqDyWipuhN6tj8Jc4EN9 cEhGo4+QcjvtisKJu84PjMjqV8d/fBfAaf6FiTtRHJtQ+qFttV7UfWQIn9pqD1iijKObAB2m FHQrJGjM6kN3xSBtn5zTzwAAAAAAAA== --------------ms080601050104070807090104-- --===============1296556000== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1296556000==-- From libssh2-devel-bounces@cool.haxx.se Tue May 24 08:40:48 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4O6eGOe009424; Tue, 24 May 2011 08:40:43 +0200 Received: from mail-pz0-f54.google.com (mail-pz0-f54.google.com [209.85.210.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4O6eDcG009174 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 08:40:14 +0200 Received: by pzk27 with SMTP id 27so3512935pzk.41 for ; Mon, 23 May 2011 23:40:09 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:sender:date:x-google-sender-auth :message-id:subject:from:to:content-type; bh=EiROJndN2UX50ED5KBzbzPchFLbLx86suLeMLdCA2s0=; b=SEswOD7QNQmvhTn+Bbl68TQAffO7kal0iYbEjxMbdBbUioySry0YDiPtARPitOFgY+ K+qWuZYkxIpZsAjMK0yeIYzBt52l+YQCp8PUCB6VCOlEA/D4mxvLf6nAt1VIrpzYvviB LBxjUVW4N1gJltbeUhljrrJKys1xLkk5KGKT4= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:date:x-google-sender-auth:message-id:subject :from:to:content-type; b=nARPMWVlC4LkseK043SkT2r6vxjb3kXX4LRXwBH6S0vuTgwgmp/1zOFM3SrlnnWErl yijUhKY8JGirqzTJsm6gkpZPwSsuyI/MWOWNC1aFGduFt1afDaffoFSSZPwd0R23Mviz 1gL3DNzfCYC6Dd+CSyrhIXd1EfXlye7QWYAX4= MIME-Version: 1.0 Received: by 10.142.180.12 with SMTP id c12mr981000wff.298.1306219209075; Mon, 23 May 2011 23:40:09 -0700 (PDT) Received: by 10.143.5.12 with HTTP; Mon, 23 May 2011 23:40:09 -0700 (PDT) Date: Tue, 24 May 2011 16:40:09 +1000 X-Google-Sender-Auth: ud-BjUXKkRAGHlb9Qrb-h61GHCI Message-ID: Subject: configure with non standard openssl installation path From: Bill Segall To: libssh2-devel@cool.haxx.se X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 08:40:48 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 08:40:15 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1335169573==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1335169573== Content-Type: multipart/alternative; boundary=000e0cd15178f4acae04a3ffdbec --000e0cd15178f4acae04a3ffdbec Content-Type: text/plain; charset=ISO-8859-1 I'm endeavouring to cross-compile libssh2-1.2.8 but I don't see a valid mechanism for telling configure the location of my openssl libraries. configure.ac seems to understand a --with-openssl which tells it that I should use that over gcrypt and it alludes to a --wtih-libssl-prefix= option but that seems not to actually have been implemented. As I'm cross-compiling libssh isn't going to be found in my system path. I would normally have reported this as a bug in your system but selecting google openid from http://trac.libssh2.org/openidlogin reports: Error in discovery: Error fetching XRDS document: (60, 'SSL certificate problem, verify that the CA cert is OK. Details:\nerror:14090086:SSL routines:SSL3_GET_SERVER_CERTIFICATE:certificate verify failed') Thanks, Bill. --000e0cd15178f4acae04a3ffdbec Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable I'm endeavouring to cross-compile libssh2-1.2.8 but I don't see a v= alid mechanism for telling configure the location of my openssl libraries.<= br>
configure.ac seems to understand= a --with-openssl which tells it that I should use that over gcrypt and it = alludes to a --wtih-libssl-prefix=3D option but that seems not to actually = have been implemented. As I'm cross-compiling libssh isn't going to= be found in my system path.

I would normally have reported this as a bug in your system but selecti= ng google openid from http:= //trac.libssh2.org/openidlogin reports:

Error in discovery: Erro= r fetching XRDS document: (60, 'SSL certificate=20 problem, verify that the CA cert is OK. Details:\nerror:14090086:SSL=20 routines:SSL3_GET_SERVER_CERTIFICATE:certificate verify failed')
Thanks,
Bill.
--000e0cd15178f4acae04a3ffdbec-- --===============1335169573== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1335169573==-- From libssh2-devel-bounces@cool.haxx.se Tue May 24 09:22:21 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4O7MFtK004999; Tue, 24 May 2011 09:22:20 +0200 Received: from yxa-v.extundo.com (yxa-v.extundo.com [213.115.69.139]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4O7MEk0004971 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NOT) for ; Tue, 24 May 2011 09:22:14 +0200 Received: from latte.josefsson.org (c80-216-4-108.bredband.comhem.se [80.216.4.108]) (authenticated bits=0) by yxa-v.extundo.com (8.14.3/8.14.3/Debian-5+lenny1) with ESMTP id p4O7M8VT011963 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES128-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 09:22:10 +0200 From: Simon Josefsson To: libssh2 development Subject: Re: configure with non standard openssl installation path References: OpenPGP: id=B565716F; url=http://josefsson.org/key.txt X-Hashcash: 1:22:110524:libssh2-devel@cool.haxx.se::0bl9zComuZtrzDY0:6Zml Date: Tue, 24 May 2011 09:22:08 +0200 In-Reply-To: (Bill Segall's message of "Tue, 24 May 2011 16:40:09 +1000") Message-ID: <877h9gsgv3.fsf@latte.josefsson.org> User-Agent: Gnus/5.110018 (No Gnus v0.18) Emacs/23.2 (gnu/linux) MIME-Version: 1.0 X-Spam-Status: No, score=-0.8 required=5.0 tests=AWL,BAYES_00, DATE_IN_FUTURE_96_XX,RDNS_DYNAMIC,SPF_FAIL autolearn=no version=3.2.5 X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on yxa-v.extundo.com X-Virus-Scanned: clamav-milter 0.97 at yxa-v X-Virus-Status: Clean X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 09:22:21 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 09:22:14 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Bill Segall writes: > I'm endeavouring to cross-compile libssh2-1.2.8 but I don't see a valid > mechanism for telling configure the location of my openssl libraries. > > configure.ac seems to understand a --with-openssl which tells it that I > should use that over gcrypt and it alludes to a --wtih-libssl-prefix= option > but that seems not to actually have been implemented. As I'm cross-compiling > libssh isn't going to be found in my system path. It is implemented. What are the commands you are using, and what error do you get? Normally you either use ./configure CPPFLAGS=foo LDFLAGS=bar to hard code library settings, but you can also use ./configure --with-openssl --with-libssl-prefix=/foo/bar/root/usr to do it in a platform independent way. /Simon _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 24 12:25:46 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OAP1xZ020263; Tue, 24 May 2011 12:25:42 +0200 Received: from mail-pz0-f54.google.com (mail-pz0-f54.google.com [209.85.210.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OAOwhg020207 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 12:24:59 +0200 Received: by pzk27 with SMTP id 27so3580521pzk.41 for ; Tue, 24 May 2011 03:24:54 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:content-type; bh=lxzJ6hE7pJ/9bU2Lm56x0wCtLjUlNvPHzJB/HhGhlcM=; b=vCRxdnTTsDmrJa7Rtxol7Ac4utgpiC72vFwBfJYHL5igemHzgpSLKBIHrXRZiGPiBc AK3DUFAiPwXA8w5eXgQEDJZ8APxz+uyupOJHxdEr6HQ5+18Z6ng7Icv1amA6xicupN9R kdk/fi+ehLia+NUNPA0rciwlqILSCmwMRrSqI= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:content-type; b=qKJA3YjGLNc9176yPg3uCn8RoYv9ZvmhiuRTWWBTSO7IpSnPO7EtHtBFA+btZXxW/3 e58tpXm7uJ3h5hj6tGXyXJ8odZ7Ma11XtPjgphp101nUUA060q8fvKiDiPG79eU5OB6y V8yZAb5Y8Ugn3iNz7bOZo1jAVMSqsqgYsd+Jo= MIME-Version: 1.0 Received: by 10.142.193.11 with SMTP id q11mr988223wff.355.1306232693304; Tue, 24 May 2011 03:24:53 -0700 (PDT) Received: by 10.143.5.12 with HTTP; Tue, 24 May 2011 03:24:53 -0700 (PDT) In-Reply-To: <877h9gsgv3.fsf@latte.josefsson.org> References: <877h9gsgv3.fsf@latte.josefsson.org> Date: Tue, 24 May 2011 20:24:53 +1000 X-Google-Sender-Auth: Vt15Zg91rI-tzDA17qjFflQw71A Message-ID: Subject: Re: configure with non standard openssl installation path From: Bill Segall To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 12:25:46 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 12:25:00 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0219609173==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============0219609173== Content-Type: multipart/alternative; boundary=000e0cd1582eadad9004a402ff55 --000e0cd1582eadad9004a402ff55 Content-Type: text/plain; charset=ISO-8859-1 My configure line is: ./configure --build=x86_64-redhat-linux --host=mips64-octeon-linux-gnu --target=mips64-octeon-linux-gnu --prefix=/site/platforms/mips64-octeon-linux/stow/libssh2-1.2.8 --with-libssl-prefix=/site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r --with-openssl --enable-static --disable-shared This reports: checking for libssl... no checking for libgcrypt... no configure: error: cannot find OpenSSL or Libgcrypt, try --with-libssl-prefix=PATH or --with-libgcrypt-prefix=PATH That's the right path to my openssl build: $ ls /site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r bin/ include/ lib/ openssl/ config.log shows: configure:12975: /site/platforms/mips64-octeon-linux/ge-cavium-linux-sdk/OCTEON-SDK/tools-gcc-4.3/bin/mips64-octeon-linux-gnu-gcc -o conftest -ggdb3 -O0 -Wall -Wextra conftest.c -lssl -lcrypto >&5 conftest.c:25:25: error: openssl/ssl.h: No such file or directory configure:12975: $? = 1 Looking at my configure.ac I see no sign of an AC_ARG_WITH(libssl-prefix) Thanks for any help, Bill. On Tue, May 24, 2011 at 5:22 PM, Simon Josefsson wrote: > Bill Segall writes: > > > I'm endeavouring to cross-compile libssh2-1.2.8 but I don't see a valid > > mechanism for telling configure the location of my openssl libraries. > > > > configure.ac seems to understand a --with-openssl which tells it that I > > should use that over gcrypt and it alludes to a --wtih-libssl-prefix= > option > > but that seems not to actually have been implemented. As I'm > cross-compiling > > libssh isn't going to be found in my system path. > > It is implemented. What are the commands you are using, and what error > do you get? > > Normally you either use ./configure CPPFLAGS=foo LDFLAGS=bar to hard > code library settings, but you can also use ./configure --with-openssl > --with-libssl-prefix=/foo/bar/root/usr to do it in a platform > independent way. > > /Simon > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > --000e0cd1582eadad9004a402ff55 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable My configure line is:
./configure --build=3Dx86_64-redhat-linux --host= =3Dmips64-octeon-linux-gnu --target=3Dmips64-octeon-linux-gnu --prefix=3D/s= ite/platforms/mips64-octeon-linux/stow/libssh2-1.2.8 --with-libssl-prefix= =3D/site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r --with-openssl -= -enable-static --disable-shared

This reports:
checking for libssl... no
checking for libgcrypt...= no
configure: error: cannot find OpenSSL or Libgcrypt,
try --with-li= bssl-prefix=3DPATH or --with-libgcrypt-prefix=3DPATH

That's the = right path to my openssl build:

=A0$ ls /site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r
bin/= =A0 include/=A0 lib/=A0 openssl/

config.log shows:
configure:1297= 5: /site/platforms/mips64-octeon-linux/ge-cavium-linux-sdk/OCTEON-SDK/tools= -gcc-4.3/bin/mips64-octeon-linux-gnu-gcc -o conftest -ggdb3 -O0 -Wall -Wext= ra=A0=A0 conftest.c=A0 -lssl -lcrypto >&5
conftest.c:25:25: error: openssl/ssl.h: No such file or directory
config= ure:12975: $? =3D 1


Looking at my configure.ac I see no sign of an AC_ARG_WITH(libssl-prefix)

Th= anks for any help,
Bill.

On Tue, May 24, 2011 at 5:22 PM, Si= mon Josefsson <= simon@josefsson.org> wrote:
Bill Segall <bill@s= egall.net> writes:

> I'm endeavouring to cross-compile libssh2-1.2.8 but I don't se= e a valid
> mechanism for telling configure the location of my openssl libraries.<= br> >
> configure.ac see= ms to understand a --with-openssl which tells it that I
> should use that over gcrypt and it alludes to a --wtih-libssl-prefix= =3D option
> but that seems not to actually have been implemented. As I'm cross= -compiling
> libssh isn't going to be found in my system path.

It is implemented. =A0What are the commands you are using, and what e= rror
do you get?

Normally you either use ./configure CPPFLAGS=3Dfoo LDFLAGS=3Dbar to hard code library settings, but you can also use ./configure --with-openssl
--with-libssl-prefix=3D/foo/bar/root/usr to do it in a platform
independent way.

/Simon
_______________________________________________
libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/li= bssh2-devel

--000e0cd1582eadad9004a402ff55-- --===============0219609173== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============0219609173==-- From libssh2-devel-bounces@cool.haxx.se Tue May 24 12:39:16 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OAdA9n026264; Tue, 24 May 2011 12:39:15 +0200 Received: from mail-pw0-f54.google.com (mail-pw0-f54.google.com [209.85.160.54]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OAd7lM026170 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 12:39:08 +0200 Received: by pwj8 with SMTP id 8so3631647pwj.41 for ; Tue, 24 May 2011 03:39:04 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:sender:date:x-google-sender-auth :message-id:subject:from:to:content-type; bh=unEaoIGDET91BzMRKpjjRpv35I2fFnPS0uIqbonGEMs=; b=Ufjb5jfIQjfrv+8rxYx4pxCZzcvH2IVue14gpIlYvjD5QsSjnkFJfGaIRBfMKpZNYC 0EUSFqhhDZKpv0k3I+Dia8w4kLFLLfnZAxJU94dnfB/tj50pRJ7DovJgkd+tP9SQpeZj rpYe5OEKSWv2iilYGcv+DO6UQx5pz+wlM3wgw= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:date:x-google-sender-auth:message-id:subject :from:to:content-type; b=mDDuNZpFfa8sQUSzwVTyLzoVhNawldYS0CJZ4iyNR70GUklAl1o6UFJcFWldXkt6bG ClzYOUadgbJyfR8v5sSIbdeQuIR5bDOrGoh5egMs1/IE320h5zDhjNTsZmdzWD5fi9Al ojWNZzKiDfUeny5fD2c4YzzqXoFUF6lSAXVG0= MIME-Version: 1.0 Received: by 10.142.208.1 with SMTP id f1mr1059620wfg.412.1306233544741; Tue, 24 May 2011 03:39:04 -0700 (PDT) Received: by 10.143.5.12 with HTTP; Tue, 24 May 2011 03:39:04 -0700 (PDT) Date: Tue, 24 May 2011 20:39:04 +1000 X-Google-Sender-Auth: FZjY4JURgAsBDy2ryKsrvepfz4Y Message-ID: Subject: A question about windows static library linkage From: Bill Segall To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 12:39:16 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 12:39:09 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0443876450==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============0443876450== Content-Type: multipart/alternative; boundary=000e0cd32c446d954004a403328e --000e0cd32c446d954004a403328e Content-Type: text/plain; charset=ISO-8859-1 Sorry to be asking so many questions today but I'm also endeavoring to build libssh2 for Windows. I want to use static libraries and to support MD, MDd, MT, and MTd variants on both win32 and x64. (Don't ask!) We're using Visual Studio9.1 and I was able to convert the VS6 dsw without any issues, cut it down to just the win32/MD[d] build and with some tweaks get it to build the library for me without any problems. When building the sample against the static library I get libssh2 undefined symbols of the form: Error 1 error LNK2001: unresolved external symbol __imp__libssh2_exit simple.obj tests which is the result of not wanting the declspec(dllimport) definitinition in libssh2.h to be active. I can resolve this by defining LIBSSH2_API in sample.c but I wanted to check that was the correct usage of that in an application. It struck me that this preprocessor symbol had the appearance of being internal to the library and I thought I should ask whether this is the defined mechanism for achieving static linkage on Windows. Thanks again, Bill. PS. I'd be happy to provide a VS9.1 solution once it was complete if that was useful. --000e0cd32c446d954004a403328e Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Sorry to be asking so many questions today but I'm also endeavoring to = build libssh2 for Windows.

I want to use static libraries and to sup= port MD, MDd, MT, and MTd variants on both win32 and x64. (Don't ask!) = We're using Visual Studio9.1 and I was able to convert the VS6 dsw with= out any issues, cut it down to just the win32/MD[d] build and with some twe= aks get it to build the library for me without any problems.

When building the sample against the static library I get libssh2 undef= ined symbols of the form:
Error=A0=A0=A0 1=A0=A0=A0 error LNK2001: unres= olved external symbol __imp__libssh2_exit=A0=A0=A0 simple.obj=A0=A0=A0 test= s

which is the result of not wanting the declspec(dllimport) definit= inition in libssh2.h to be active. I can resolve this by defining LIBSSH2_A= PI in sample.c but I wanted to check that was the correct usage of that in = an application. It struck me that this preprocessor symbol had the appearan= ce of being internal to the library and I thought I should ask whether this= is the defined mechanism for achieving static linkage on Windows.

Thanks again,
Bill.
PS. I'd be happy to provide a VS9.1 solut= ion once it was complete if that was useful.


--000e0cd32c446d954004a403328e-- --===============0443876450== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============0443876450==-- From libssh2-devel-bounces@cool.haxx.se Tue May 24 13:41:01 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OBejJt000490; Tue, 24 May 2011 13:40:59 +0200 Received: from yxa-v.extundo.com (yxa-v.extundo.com [213.115.69.139]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OBegtd000428 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NOT) for ; Tue, 24 May 2011 13:40:43 +0200 Received: from latte.josefsson.org (host-90-233-172-47.mobileonline.telia.com [90.233.172.47]) (authenticated bits=0) by yxa-v.extundo.com (8.14.3/8.14.3/Debian-5+lenny1) with ESMTP id p4OBeLdU024081 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES128-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 13:40:30 +0200 From: Simon Josefsson To: libssh2 development Subject: Re: configure with non standard openssl installation path References: <877h9gsgv3.fsf@latte.josefsson.org> OpenPGP: id=B565716F; url=http://josefsson.org/key.txt X-Hashcash: 1:22:110524:libssh2-devel@cool.haxx.se::Ym9zJ5MZmc7dDiqX:N7B2 Date: Tue, 24 May 2011 13:40:20 +0200 Message-ID: <874o4knx7f.fsf@latte.josefsson.org> MIME-Version: 1.0 X-Spam-Status: No, score=-0.4 required=5.0 tests=BAYES_00, DATE_IN_FUTURE_96_XX, RDNS_DYNAMIC,SPF_FAIL autolearn=no version=3.2.5 X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on yxa-v.extundo.com X-Virus-Scanned: clamav-milter 0.97 at yxa-v X-Virus-Status: Clean X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 13:41:01 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 13:40:43 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Bill Segall writes: > My configure line is: > ./configure --build=x86_64-redhat-linux --host=mips64-octeon-linux-gnu > --target=mips64-octeon-linux-gnu > --prefix=/site/platforms/mips64-octeon-linux/stow/libssh2-1.2.8 > --with-libssl-prefix=/site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r > --with-openssl --enable-static --disable-shared > > This reports: > checking for libssl... no > checking for libgcrypt... no > configure: error: cannot find OpenSSL or Libgcrypt, > try --with-libssl-prefix=PATH or --with-libgcrypt-prefix=PATH > > That's the right path to my openssl build: > > $ ls /site/platforms/mips64-octeon-linux/stow/openssl-0.9.8r > bin/ include/ lib/ openssl/ > > config.log shows: > configure:12975: > /site/platforms/mips64-octeon-linux/ge-cavium-linux-sdk/OCTEON-SDK/tools-gcc-4.3/bin/mips64-octeon-linux-gnu-gcc > -o conftest -ggdb3 -O0 -Wall -Wextra conftest.c -lssl -lcrypto >&5 > conftest.c:25:25: error: openssl/ssl.h: No such file or directory > configure:12975: $? = 1 Isn't there another attempt after that? I think the code tries without flags first, and then tries to add them to see if they are really needed. > Looking at my configure.ac I see no sign of an > AC_ARG_WITH(libssl-prefix) Check m4/lib*.m4. /Simon > > Thanks for any help, > Bill. > > On Tue, May 24, 2011 at 5:22 PM, Simon Josefsson wrote: > >> Bill Segall writes: >> >> > I'm endeavouring to cross-compile libssh2-1.2.8 but I don't see a valid >> > mechanism for telling configure the location of my openssl libraries. >> > >> > configure.ac seems to understand a --with-openssl which tells it that I >> > should use that over gcrypt and it alludes to a --wtih-libssl-prefix= >> option >> > but that seems not to actually have been implemented. As I'm >> cross-compiling >> > libssh isn't going to be found in my system path. >> >> It is implemented. What are the commands you are using, and what error >> do you get? >> >> Normally you either use ./configure CPPFLAGS=foo LDFLAGS=bar to hard >> code library settings, but you can also use ./configure --with-openssl >> --with-libssl-prefix=/foo/bar/root/usr to do it in a platform >> independent way. >> >> /Simon >> _______________________________________________ >> libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel >> > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 24 13:43:20 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OBhHsg003301; Tue, 24 May 2011 13:43:18 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OBhG5V003271 for ; Tue, 24 May 2011 13:43:16 +0200 Date: Tue, 24 May 2011 13:43:16 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: configure with non standard openssl installation path In-Reply-To: <874o4knx7f.fsf@latte.josefsson.org> Message-ID: References: <877h9gsgv3.fsf@latte.josefsson.org> <874o4knx7f.fsf@latte.josefsson.org> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 13:43:20 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 13:43:16 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 24 May 2011, Simon Josefsson wrote: >> Looking at my configure.ac I see no sign of an >> AC_ARG_WITH(libssl-prefix) > > Check m4/lib*.m4. It is slightly obfuscated. See m4/lib-link.m4 around line 157. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 24 14:19:36 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OCJU1o024171; Tue, 24 May 2011 14:19:35 +0200 Received: from mail-px0-f177.google.com (mail-px0-f177.google.com [209.85.212.177]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4OCJSI7024152 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 24 May 2011 14:19:29 +0200 Received: by pxi10 with SMTP id 10so4725569pxi.8 for ; Tue, 24 May 2011 05:19:23 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:content-type; bh=tslH44zJ/j6ELcksPhAWBVMM/MaE0OB2N+0WNTTYg1s=; b=nHbUZOCbiyD+lasAnyCTin2hnCXEbaUU+BAfVGGqnw5RCfA2rzNmhlRhudWfiEa+QU Gtn6TauVmCOOc4+Z2uRXRq4K+/WmjWrgCTpU3kv9D1Wzaw4CQFAAaMEVM+Vk4dFJdDvX Ro3eJ/C1BxJHJWbWAOycwrlnrLhqJdLrtHzQA= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:content-type; b=ngBeDBPTKCeUWRmxk7E4sfwz7VYPWJBp/c813KxtAlju5hZKgkmtsCFGqNYwfIHDiA xDM6/QTELLxD6ecti4ZtfAYnszWVEI2ECC6vVGSv/f/tVeOMjP4+AJE4Q6wY8UrGwNdb 5Bqt4J9s/pHzLD2pmnt9vZgLtcJne1iozSqfc= MIME-Version: 1.0 Received: by 10.142.193.11 with SMTP id q11mr1000996wff.355.1306239563858; Tue, 24 May 2011 05:19:23 -0700 (PDT) Received: by 10.143.5.12 with HTTP; Tue, 24 May 2011 05:19:23 -0700 (PDT) In-Reply-To: References: <877h9gsgv3.fsf@latte.josefsson.org> <874o4knx7f.fsf@latte.josefsson.org> Date: Tue, 24 May 2011 22:19:23 +1000 X-Google-Sender-Auth: CExTKif-oiaSOv5NITYcWB9MsVY Message-ID: Subject: Re: configure with non standard openssl installation path From: Bill Segall To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 24 May 2011 14:19:36 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 24 May 2011 14:19:30 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1889456987==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1889456987== Content-Type: multipart/alternative; boundary=000e0cd1582e3202e804a4049947 --000e0cd1582e3202e804a4049947 Content-Type: text/plain; charset=ISO-8859-1 Thanks, that was the hint I needed. By adding a set -x so I could see what the shell was doing, I was able to determine that it was looking in $withval/lib64 and by linking lib64 to lib I can get configure to run successfully and appear sane. I'm not quite of the woods yet with my libtool but that looks like I need to specify some additional cross compile tools for my libtool. Many thanks for your assistance. Bill. On Tue, May 24, 2011 at 9:43 PM, Daniel Stenberg wrote: > On Tue, 24 May 2011, Simon Josefsson wrote: > > Looking at my configure.ac I see no sign of an >>> AC_ARG_WITH(libssl-prefix) >>> >> >> Check m4/lib*.m4. >> > > It is slightly obfuscated. See m4/lib-link.m4 around line 157. > > -- > > / daniel.haxx.se > > _______________________________________________ > libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel > --000e0cd1582e3202e804a4049947 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Thanks, that was the hint I needed. By adding a set -x so I could see what = the shell was doing, I was able to determine that it was looking in $withva= l/lib64 and by linking lib64 to lib I can get configure to run successfully= and appear sane.

I'm not quite of the woods yet with my libtool but that looks like = I need to specify some additional cross compile tools for my libtool.
Many thanks for your assistance.

Bill.

On Tue, May 24, 2011 at 9:43 PM, Daniel Stenberg <daniel@haxx.se> wrote:
On Tue, 24 May 2011, Simon Josefsson wrote:

Looking at my configure.a= c I see no sign of an
AC_ARG_WITH(libssl-prefix)

Check m4/lib*.m4.

It is slightly obfuscated. See m4/lib-link.m4 around line 157.

--

=A0/ daniel.haxx.se=

_______________________________________________
libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/li= bssh2-devel

--000e0cd1582e3202e804a4049947-- --===============1889456987== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1889456987==-- From libssh2-devel-bounces@cool.haxx.se Wed May 25 03:00:00 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4P0xdGZ021136; Wed, 25 May 2011 02:59:57 +0200 Received: from mail102.csoft.net (postfix@mail102.csoft.net [205.205.219.4]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4P0xclo021133 for ; Wed, 25 May 2011 02:59:38 +0200 Received: from mail102.csoft.net (localhost [127.0.0.1]) by mail102.csoft.net (Postfix) with ESMTP id B770F3EF1771 for ; Tue, 24 May 2011 20:59:45 -0400 (EDT) DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=csoft.net; h=message-id :date:from:mime-version:to:subject:content-type :content-transfer-encoding; s=sel1; bh=To7981ohfRr52j/k5mZ/O/juK Lw=; b=R1DyG9jOVgOTYY5Gnq6dDyhbMK6nF6iuDPSp7OjFm4qWgnGBeyTTZxazU XN92QWTgAW5IxOWxdP76cehWjbP6I81zidgQMaQB1VKLGNebCXvgt3BWL/eIvWok z1tH67jLcsGhDk6UthnyrdMPypb8y8fh3/p7YznZZDsLuTQARY= Received: from [172.28.133.90] (natint3.juniper.net [66.129.224.36]) (using TLSv1 with cipher DHE-RSA-CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) by mail102.csoft.net (Postfix) with ESMTPSA id 5198C3EF0856 for ; Tue, 24 May 2011 20:59:45 -0400 (EDT) Message-ID: <4DDC5476.3010309@fid4.com> Date: Tue, 24 May 2011 20:59:34 -0400 From: "Michael C. Cambria" User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.17) Gecko/20110414 Thunderbird/3.1.10 MIME-Version: 1.0 To: libssh2-devel@cool.haxx.se Subject: Starting a subsystem other than e.g. sftp X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 25 May 2011 03:00:00 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 25 May 2011 02:59:38 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se Hi, Is there a way to supply to libssh2 the name of an external subsystem which sshd is to use (similar to "-s my-subsystem" on the command line in OpenSSH?) Thanks, MikeC _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Wed May 25 16:20:46 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4PEKLS9018881; Wed, 25 May 2011 16:20:42 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4PEKJxI018838 for ; Wed, 25 May 2011 16:20:19 +0200 Date: Wed, 25 May 2011 16:20:19 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Starting a subsystem other than e.g. sftp In-Reply-To: <4DDC5476.3010309@fid4.com> Message-ID: References: <4DDC5476.3010309@fid4.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Wed, 25 May 2011 16:20:46 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Wed, 25 May 2011 16:20:19 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 24 May 2011, Michael C. Cambria wrote: > Is there a way to supply to libssh2 the name of an external subsystem which > sshd is to use (similar to "-s my-subsystem" on the command line in > OpenSSH?) You can start a custom subsystem with libssh2_channel_process_startup(). Something like: libssh2_channel_process_startup(handle, "subsystem", 9, "my-subsystem", 12); You can probably check out the sftp.c code to get inspiration, since SFTP is a subsystem. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Sun May 29 21:54:31 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4TJs5to007803; Sun, 29 May 2011 21:54:27 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4TJs2RF007716 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Sun, 29 May 2011 21:54:04 +0200 Received: by pvg11 with SMTP id 11so1729667pvg.41 for ; Sun, 29 May 2011 12:53:59 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:date:message-id:subject:from:to :content-type; bh=exAwPfoBf6i1Vt480GY9JJZhxfe724cg+CRI9nM3YT8=; b=RF2q0OPP31nvMIEpy9uy8UCFapP8dxnCGiwI2rLtu4RlJy7YiPyOuePfwvjHYvNHrZ hvRS9Nwe/Shf9Uk9REVwcTTiz3W8GSbwMSQ1B1qvWv2PipN1t1wUNrZVEoaqbxaJTAxZ Z8/BdJXPX+lPjFaVQROtiLrpEafcnCruscNw0= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=mkyqI7t4VWJNwXZ0o/wH8/Tsf4F3X4A+iHcQSvfvFyiKJbXfrKj1oEKT6IUOf6Wxmx gscxlakVQDHTrpH3sue1JJCPFGIP970OHo+KwGaNLdE53Ze2sVvD1A1dw+r7HdgdT6DZ owhopHNG17FGEwJ2uq98EqaWrojym6uvAno2c= MIME-Version: 1.0 Received: by 10.68.17.99 with SMTP id n3mr1733013pbd.351.1306698839612; Sun, 29 May 2011 12:53:59 -0700 (PDT) Received: by 10.68.71.165 with HTTP; Sun, 29 May 2011 12:53:59 -0700 (PDT) Date: Sun, 29 May 2011 15:53:59 -0400 Message-ID: Subject: Core dump when authentication fails? From: Saqib Ali To: libssh2-devel@cool.haxx.se X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Sun, 29 May 2011 21:54:31 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Sun, 29 May 2011 21:54:04 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============1085982220==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============1085982220== Content-Type: multipart/alternative; boundary=bcaec52161f129dc4e04a46f881f --bcaec52161f129dc4e04a46f881f Content-Type: text/plain; charset=ISO-8859-1 Cross posted from: http://curl.haxx.se/mail/lib-2011-05/0209.html I'm using Curl version 7.21.5 to do SFTP uploads from a Solaris 10 box. It works fine for me when I give it valid parameters (username, password, IP address) However, I'm testing it in the exact same scenario -- only with the wrong password. I'm expecting that it will be stopped because the transfer is interrupted by the timeout duration.. But instead of that, I'm getting a core dump. (See below) Has anyone else seen this behavior? What is the correct expected behavior when the user authentication fails? - Saqib * About to connect() to xxx.xxx.xxx.xxx port 22 (#0) * Trying xxx.xxx.xxx.xxx... * 0xbce30 is at send pipe head! * STATE: CONNECT => WAITCONNECT handle 0x93cb0; (connection #0) * Connected to xxx.xxx.xxx.xxx (xxx.xxx.xxx.xxx) port 22 (#0) * SFTP 0xdcc38 state change from SSH_STOP to SSH_S_STARTUP * SFTP 0xdcc38 state change from SSH_S_STARTUP to SSH_HOSTKEY * SFTP 0xdcc38 state change from SSH_HOSTKEY to SSH_AUTHLIST * STATE: WAITCONNECT => PROTOCONNECT handle 0x93cb0; (connection #0) * SSH authentication methods available: gssapi-keyex,gssapi-with-mic,publickey,password,keyboard-interactive * SFTP 0xdcc38 state change from SSH_AUTHLIST to SSH_AUTH_PKEY_INIT * Using ssh public key file /SS8/users/lis/.ssh/id_dsa.pub * Using ssh private key file /SS8/users/lis/.ssh/id_dsa * SFTP 0xdcc38 state change from SSH_AUTH_PKEY_INIT to SSH_AUTH_PKEY * SSH public key authentication failed: Unable to open public key file * SFTP 0xdcc38 state change from SSH_AUTH_PKEY to SSH_AUTH_PASS_INIT * SFTP 0xdcc38 state change from SSH_AUTH_PASS_INIT to SSH_AUTH_PASS * SFTP 0xdcc38 state change from SSH_AUTH_PASS to SSH_AUTH_HOST_INIT * SFTP 0xdcc38 state change from SSH_AUTH_HOST_INIT to SSH_AUTH_KEY_INIT * SFTP 0xdcc38 state change from SSH_AUTH_KEY_INIT to SSH_AUTH_KEY Assertion failed: wantedsize != 0, file memdebug.c, line 136 Abort (core dumped) - Saqib --bcaec52161f129dc4e04a46f881f Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable



- Sa= qib


--bcaec52161f129dc4e04a46f881f-- --===============1085982220== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============1085982220==-- From libssh2-devel-bounces@cool.haxx.se Sun May 29 23:21:14 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4TLKwDF016868; Sun, 29 May 2011 23:21:12 +0200 Received: from mail102.csoft.net (postfix@mail102.csoft.net [205.205.219.4]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4TLKuTX016846 for ; Sun, 29 May 2011 23:20:56 +0200 Received: from mail102.csoft.net (localhost [127.0.0.1]) by mail102.csoft.net (Postfix) with ESMTP id 9C2853EF174F for ; Sun, 29 May 2011 17:20:56 -0400 (EDT) DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=csoft.net; h=message-id :date:from:mime-version:to:subject:references:in-reply-to :content-type:content-transfer-encoding; s=sel1; bh=2W4XZBZb9v28 sMXD9e2cig+j7bU=; b=aSqzI5nPK+MInhAX/T1URVurdAf7YNDkUXLUVQxFJh6O 3GUK+imPPNtmYj92CIbB8sSR6tlkGqXaBrXWTJQSKc5KXkVKLgXwrO6UEDpmJSWn RAtl/ZiI9SjBZWreyVlcwU5ZrerBnI1YAb4pR8+5VwMY9gUZJJ/qI3BupZPYIvA= Received: from [172.16.8.69] (c-24-34-109-241.hsd1.ma.comcast.net [24.34.109.241]) (using TLSv1 with cipher DHE-RSA-CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) by mail102.csoft.net (Postfix) with ESMTPSA id 7395A3EF174E for ; Sun, 29 May 2011 17:20:56 -0400 (EDT) Message-ID: <4DE2B8B1.9050101@fid4.com> Date: Sun, 29 May 2011 17:20:49 -0400 From: "Michael C. Cambria" User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.17) Gecko/20110414 Thunderbird/3.1.10 MIME-Version: 1.0 To: libssh2 development Subject: Re: Starting a subsystem other than e.g. sftp References: <4DDC5476.3010309@fid4.com> In-Reply-To: <4DDC5476.3010309@fid4.com> X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Sun, 29 May 2011 23:21:14 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Sun, 29 May 2011 23:20:56 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On 5/24/2011 8:59 PM, Michael C. Cambria wrote: > > Hi, > > Is there a way to supply to libssh2 the name of an external subsystem > which sshd is to use (similar to "-s my-subsystem" on the command line > in OpenSSH?) > libssh2_channel_subsystem(channel, commandline) seems to work for me as well _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Mon May 30 10:35:31 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4U8Z9Y7008921; Mon, 30 May 2011 10:35:28 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4U8Z7MJ008911 for ; Mon, 30 May 2011 10:35:07 +0200 Date: Mon, 30 May 2011 10:35:06 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Core dump when authentication fails? In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Mon, 30 May 2011 10:35:31 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Mon, 30 May 2011 10:35:07 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Sun, 29 May 2011, Saqib Ali wrote: > Cross posted from: http://curl.haxx.se/mail/lib-2011-05/0209.html Allow me to it-iterate here what I mention on the libcurl list: that check is libcurl's way to detect problems in libcurl and not necessarily an error when made by "others" such as libssh2. Also, the assert() you see in libcurl is only present when you build libcurl debug-enabled and thus it will behave differently if you build a "normal" build. I do however think it is a REALLY bad idea to malloc() zero bytes (because we can't rely on the return code) so we should correct libssh2 to not do that. I guess the code for the keyboard-interactive auth must not assume that the data that arrives actually has non-zero lengths. > Has anyone else seen this behavior? What is the correct expected behavior > when the user authentication fails? Please don't assume that people on this list knows about libcurl or cares about it to any particular extent. This list is about libssh2. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 31 23:38:08 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLbjgb007572; Tue, 31 May 2011 23:38:04 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLbiGt007564 for ; Tue, 31 May 2011 23:37:44 +0200 Date: Tue, 31 May 2011 23:37:44 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: timeouts in libssh2 blocking-mode In-Reply-To: <4DC07E17.7080106@securitease.com> Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> <4DC06BDA.2040605@securitease.com> <4DC07E17.7080106@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 31 May 2011 23:38:08 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 31 May 2011 23:37:44 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Wed, 4 May 2011, Matt Lilley wrote: > I think this patch does that - but it's early in the morning still. Let me > know what you think It just struck me today when browsing this code (again) that your fix completely left out the majority of all users: those who use poll()... -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 31 23:38:59 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLcv3s007842; Tue, 31 May 2011 23:38:58 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLcv6K007836 for ; Tue, 31 May 2011 23:38:57 +0200 Date: Tue, 31 May 2011 23:38:57 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Core dump when authentication fails? In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 31 May 2011 23:38:59 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 31 May 2011 23:38:57 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Mon, 30 May 2011, Daniel Stenberg wrote: > I guess the code for the keyboard-interactive auth must not assume that the > data that arrives actually has non-zero lengths. I had an attempt at fixing that code just now and pushed my fixes to the git repo. Can you please try again with the current git version and see how it works for you? -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel From libssh2-devel-bounces@cool.haxx.se Tue May 31 23:46:43 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLkewg010529; Tue, 31 May 2011 23:46:43 +0200 Received: from mail-pv0-f182.google.com (mail-pv0-f182.google.com [74.125.83.182]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLkbwK010518 (version=TLSv1/SSLv3 cipher=RC4-SHA bits=128 verify=NOT) for ; Tue, 31 May 2011 23:46:38 +0200 Received: by pvg11 with SMTP id 11so2819185pvg.41 for ; Tue, 31 May 2011 14:46:33 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=fs+1akqv38aGx7hlXIqyhGBxNYe9IEOiw0osLXDl/hI=; b=OudsyGyoD6NeJsTqjJFNu88mTANPlGFX1wqRmUKbJwrRG93gU3VVLmeIJR4wNXEcnh dUoRdh+ooHcy4lMUxdRw+m3KuRPbftQLcsy3Cm9sQJMvRmjkAKbhJrEsDFb4UnbDGVIj SkFy26GC0HJ7PAdbDaXKjaTwHH7DU3dfFc0rs= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=hyd8P3x7SaZ0l7RRoois0L2UAir8poh2u/rEi59fF82+JvnHs6/aZhsIOPqubM0Exb /OMLR7WnIWAJhlulBlRoKFVE1Ou5VMGulraNcSh7YA8oa2Uzmfpc7RAU1iPHTKzkwI6k 7p6/KBfRlKhg4+Gun1sxtLKyCxybiu+Dl799o= MIME-Version: 1.0 Received: by 10.68.17.195 with SMTP id q3mr2674614pbd.369.1306878393603; Tue, 31 May 2011 14:46:33 -0700 (PDT) Received: by 10.68.58.137 with HTTP; Tue, 31 May 2011 14:46:33 -0700 (PDT) In-Reply-To: References: Date: Tue, 31 May 2011 17:46:33 -0400 Message-ID: Subject: Re: Core dump when authentication fails? From: Saqib Ali To: libssh2 development X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 31 May 2011 23:46:43 +0200 (CEST) X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 31 May 2011 23:46:39 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0997530561==" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se --===============0997530561== Content-Type: multipart/alternative; boundary=bcaec520ea3f6a555f04a49956ff --bcaec520ea3f6a555f04a49956ff Content-Type: text/plain; charset=ISO-8859-1 On Tue, May 31, 2011 at 5:38 PM, Daniel Stenberg wrote: > On Mon, 30 May 2011, Daniel Stenberg wrote: > > I guess the code for the keyboard-interactive auth must not assume that >> the data that arrives actually has non-zero lengths. >> > > I had an attempt at fixing that code just now and pushed my fixes to the > git repo. Can you please try again with the current git version and see how > it works for you? > > > Sorry I didn't email previously about this. I didn't realize you were waiting for me. I made the patches you described in SSH2. Then I compiled CURL without the -debug-enabled flag and it did work properly. I'll still test your changes. But can I get them from the daily snapshot for June 1, 2011? --bcaec520ea3f6a555f04a49956ff Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable


On Tue, May 31, 2011 at 5:38 = PM, Daniel Stenberg <daniel@haxx.se> wrote:
On Mon, 30 May 2011, Daniel Stenberg wrote:

I guess the code for the keyboard-interactive auth must not assume that the= data that arrives actually has non-zero lengths.

I had an attempt at fixing that code just now and pushed my fixes to the gi= t repo. Can you please try again with the current git version and see how i= t works for you?



Sorry I didn't email previously about this. I didn&= #39;t realize you were waiting for me.=A0
I made the patches you = described in SSH2. Then I compiled CURL without the -debug-enabled flag and= it did work properly.

I'll still test your changes. But can I get them fr= om the daily snapshot for June 1, 2011?
--bcaec520ea3f6a555f04a49956ff-- --===============0997530561== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --===============0997530561==-- From libssh2-devel-bounces@cool.haxx.se Tue May 31 23:51:46 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLpg5V012266; Tue, 31 May 2011 23:51:45 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLpe77012259 for ; Tue, 31 May 2011 23:51:40 +0200 Date: Tue, 31 May 2011 23:51:40 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: [PATCH] Re: timeouts in libssh2 blocking-mode In-Reply-To: Message-ID: References: <4DBE40D6.9010703@securitease.com> <20110502063335.20480.qmail@stuge.se> <4DBF104B.2010508@securitease.com> <4DBF3D9D.7080800@securitease.com> <20110502235937.32306.qmail@stuge.se> <4DBF4A05.9050403@securitease.com> <20110503004825.5079.qmail@stuge.se> <4DBF5418.4040706@securitease.com> <4DC06BDA.2040605@securitease.com> <4DC07E17.7080106@securitease.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="1129329158-2020387728-1306878700=:26627" X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 31 May 2011 23:51:46 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 31 May 2011 23:51:40 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --1129329158-2020387728-1306878700=:26627 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed On Tue, 31 May 2011, Daniel Stenberg wrote: > It just struck me today when browsing this code (again) that your fix > completely left out the majority of all users: those who use poll()... Okay, I suggest a little massaging of the function like the attached patch to make the timeout work the same way when libssh2 is built to use poll(). I'm interested in a review/feedback. If nobody finds anything upsetting I'll push this in a day or two. -- / daniel.haxx.se --1129329158-2020387728-1306878700=:26627 Content-Type: TEXT/x-diff; name=0001-_libssh2_wait_socket-fix-timeouts-for-poll-uses.patch Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename=0001-_libssh2_wait_socket-fix-timeouts-for-poll-uses.patch RnJvbSBhNWRlZjUyNzZjNWI1NmRkZTVhYTAwOWNjYjYzZjdmNWQ1M2I4NWRm IE1vbiBTZXAgMTcgMDA6MDA6MDAgMjAwMQ0KRnJvbTogRGFuaWVsIFN0ZW5i ZXJnIDxkYW5pZWxAaGF4eC5zZT4NCkRhdGU6IFR1ZSwgMzEgTWF5IDIwMTEg MjM6NDk6MzIgKzAyMDANClN1YmplY3Q6IFtQQVRDSF0gX2xpYnNzaDJfd2Fp dF9zb2NrZXQ6IGZpeCB0aW1lb3V0cyBmb3IgcG9sbCgpIHVzZXMNCg0KLS0t DQogc3JjL3Nlc3Npb24uYyB8ICAxMzEgKysrKysrKysrKysrKysrKysrKysr KysrKysrKystLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tDQogMSBmaWxl cyBjaGFuZ2VkLCA2NyBpbnNlcnRpb25zKCspLCA2NCBkZWxldGlvbnMoLSkN Cg0KZGlmZiAtLWdpdCBhL3NyYy9zZXNzaW9uLmMgYi9zcmMvc2Vzc2lvbi5j DQppbmRleCBkMzcxZmYwLi4zY2Y5MmVjIDEwMDY0NA0KLS0tIGEvc3JjL3Nl c3Npb24uYw0KKysrIGIvc3JjL3Nlc3Npb24uYw0KQEAgLTU0OSw2ICs1NDks OCBAQCBpbnQgX2xpYnNzaDJfd2FpdF9zb2NrZXQoTElCU1NIMl9TRVNTSU9O ICpzZXNzaW9uLCB0aW1lX3Qgc3RhcnRfdGltZSkNCiAgICAgaW50IHNlY29u ZHNfdG9fbmV4dDsNCiAgICAgaW50IGRpcjsNCiAgICAgaW50IGhhc190aW1l b3V0Ow0KKyAgICBsb25nIG1zX3RvX25leHQgPSAwOw0KKyAgICBsb25nIGVs YXBzZWRfbXM7DQogDQogICAgIC8qIHNpbmNlIGxpYnNzaDIgb2Z0ZW4gc2V0 cyBFQUdBSU4gaW50ZXJuYWxseSBiZWZvcmUgdGhpcyBmdW5jdGlvbiBpcw0K ICAgICAgICBjYWxsZWQsIHdlIGNhbiBkZWNyZWFzZSBzb21lIGFtb3VudCBv ZiBjb25mdXNpb24gaW4gdXNlciBwcm9ncmFtcyBieQ0KQEAgLTU1OSw4MSAr NTYxLDgyIEBAIGludCBfbGlic3NoMl93YWl0X3NvY2tldChMSUJTU0gyX1NF U1NJT04gKnNlc3Npb24sIHRpbWVfdCBzdGFydF90aW1lKQ0KICAgICByYyA9 IGxpYnNzaDJfa2VlcGFsaXZlX3NlbmQgKHNlc3Npb24sICZzZWNvbmRzX3Rv X25leHQpOw0KICAgICBpZiAocmMgPCAwKQ0KICAgICAgICAgcmV0dXJuIHJj Ow0KLSAgICBlbHNlIHsNCi0gICAgICAgIC8qIGZpZ3VyZSBvdXQgd2hhdCB0 byB3YWl0IGZvciAqLw0KLSAgICAgICAgZGlyID0gbGlic3NoMl9zZXNzaW9u X2Jsb2NrX2RpcmVjdGlvbnMoc2Vzc2lvbik7DQogDQotICAgICAgICBpZigh ZGlyKSB7DQotICAgICAgICAgICAgX2xpYnNzaDJfZGVidWcoc2Vzc2lvbiwg TElCU1NIMl9UUkFDRV9TT0NLRVQsDQotICAgICAgICAgICAgICAgICAgICAg ICAgICAgIk5vdGhpbmcgdG8gd2FpdCBmb3IgaW4gd2FpdF9zb2NrZXQiKTsN Ci0gICAgICAgICAgICAvKiBUbyBhdm9pZCB0aGF0IHdlIGhhbmcgYmVsb3cg anVzdCBiZWNhdXNlIHRoZXJlJ3Mgbm90aGluZyBzZXQgdG8NCi0gICAgICAg ICAgICAgICB3YWl0IGZvciwgd2UgdGltZW91dCBvbiAxIHNlY29uZCB0byBh bHNvIGF2b2lkIGJ1c3ktbG9vcGluZw0KLSAgICAgICAgICAgICAgIGR1cmlu ZyB0aGlzIGNvbmRpdGlvbiAqLw0KLSAgICAgICAgICAgIHNlY29uZHNfdG9f bmV4dCA9IDE7DQorICAgIG1zX3RvX25leHQgPSBzZWNvbmRzX3RvX25leHQg KiAxMDAwOw0KKw0KKyAgICAvKiBmaWd1cmUgb3V0IHdoYXQgdG8gd2FpdCBm b3IgKi8NCisgICAgZGlyID0gbGlic3NoMl9zZXNzaW9uX2Jsb2NrX2RpcmVj dGlvbnMoc2Vzc2lvbik7DQorDQorICAgIGlmKCFkaXIpIHsNCisgICAgICAg IF9saWJzc2gyX2RlYnVnKHNlc3Npb24sIExJQlNTSDJfVFJBQ0VfU09DS0VU LA0KKyAgICAgICAgICAgICAgICAgICAgICAgIk5vdGhpbmcgdG8gd2FpdCBm b3IgaW4gd2FpdF9zb2NrZXQiKTsNCisgICAgICAgIC8qIFRvIGF2b2lkIHRo YXQgd2UgaGFuZyBiZWxvdyBqdXN0IGJlY2F1c2UgdGhlcmUncyBub3RoaW5n IHNldCB0bw0KKyAgICAgICAgICAgd2FpdCBmb3IsIHdlIHRpbWVvdXQgb24g MSBzZWNvbmQgdG8gYWxzbyBhdm9pZCBidXN5LWxvb3BpbmcNCisgICAgICAg ICAgIGR1cmluZyB0aGlzIGNvbmRpdGlvbiAqLw0KKyAgICAgICAgbXNfdG9f bmV4dCA9IDEwMDA7DQorICAgIH0NCisNCisgICAgaWYgKHNlc3Npb24tPmFw aV90aW1lb3V0ID4gMCAmJg0KKyAgICAgICAgKHNlY29uZHNfdG9fbmV4dCA9 PSAwIHx8DQorICAgICAgICAgc2Vjb25kc190b19uZXh0ID4gc2Vzc2lvbi0+ YXBpX3RpbWVvdXQpKSB7DQorICAgICAgICB0aW1lX3Qgbm93ID0gdGltZSAo TlVMTCk7DQorICAgICAgICBlbGFwc2VkX21zID0gKGxvbmcpKDEwMDAqZGlm ZnRpbWUoc3RhcnRfdGltZSwgbm93KSk7DQorICAgICAgICBpZiAoZWxhcHNl ZF9tcyA+IHNlc3Npb24tPmFwaV90aW1lb3V0KSB7DQorICAgICAgICAgICAg c2Vzc2lvbi0+ZXJyX2NvZGUgPSBMSUJTU0gyX0VSUk9SX1RJTUVPVVQ7DQor ICAgICAgICAgICAgcmV0dXJuIExJQlNTSDJfRVJST1JfVElNRU9VVDsNCiAg ICAgICAgIH0NCi0gICAgICAgIHsNCisgICAgICAgIG1zX3RvX25leHQgPSAo c2Vzc2lvbi0+YXBpX3RpbWVvdXQgLSBlbGFwc2VkX21zKTsNCisgICAgICAg IGhhc190aW1lb3V0ID0gMTsNCisgICAgfQ0KKyAgICBlbHNlIGlmIChtc190 b19uZXh0ID4gMCkgew0KKyAgICAgICAgaGFzX3RpbWVvdXQgPSAxOw0KKyAg ICB9DQorICAgIGVsc2UNCisgICAgICAgIGhhc190aW1lb3V0ID0gMDsNCisN CiAjaWZkZWYgSEFWRV9QT0xMDQotICAgICAgICAgICAgc3RydWN0IHBvbGxm ZCBzb2NrZXRzWzFdOw0KKyAgICB7DQorICAgICAgICBzdHJ1Y3QgcG9sbGZk IHNvY2tldHNbMV07DQogDQotICAgICAgICAgICAgc29ja2V0c1swXS5mZCA9 IHNlc3Npb24tPnNvY2tldF9mZDsNCi0gICAgICAgICAgICBzb2NrZXRzWzBd LmV2ZW50cyA9IDA7DQotICAgICAgICAgICAgc29ja2V0c1swXS5yZXZlbnRz ID0gMDsNCisgICAgICAgIHNvY2tldHNbMF0uZmQgPSBzZXNzaW9uLT5zb2Nr ZXRfZmQ7DQorICAgICAgICBzb2NrZXRzWzBdLmV2ZW50cyA9IDA7DQorICAg ICAgICBzb2NrZXRzWzBdLnJldmVudHMgPSAwOw0KIA0KLSAgICAgICAgICAg IGlmKGRpciAmIExJQlNTSDJfU0VTU0lPTl9CTE9DS19JTkJPVU5EKQ0KLSAg ICAgICAgICAgICAgICBzb2NrZXRzWzBdLmV2ZW50cyB8PSBQT0xMSU47DQor ICAgICAgICBpZihkaXIgJiBMSUJTU0gyX1NFU1NJT05fQkxPQ0tfSU5CT1VO RCkNCisgICAgICAgICAgICBzb2NrZXRzWzBdLmV2ZW50cyB8PSBQT0xMSU47 DQogDQotICAgICAgICAgICAgaWYoZGlyICYgTElCU1NIMl9TRVNTSU9OX0JM T0NLX09VVEJPVU5EKQ0KLSAgICAgICAgICAgICAgICBzb2NrZXRzWzBdLmV2 ZW50cyB8PSBQT0xMT1VUOw0KKyAgICAgICAgaWYoZGlyICYgTElCU1NIMl9T RVNTSU9OX0JMT0NLX09VVEJPVU5EKQ0KKyAgICAgICAgICAgIHNvY2tldHNb MF0uZXZlbnRzIHw9IFBPTExPVVQ7DQogDQotICAgICAgICAgICAgcmMgPSBw b2xsKHNvY2tldHMsIDEsIHNlY29uZHNfdG9fbmV4dCA/DQotICAgICAgICAg ICAgICAgICAgICAgIHNlY29uZHNfdG9fbmV4dCAqIDEwMDAgOiAtMSk7DQor ICAgICAgICByYyA9IHBvbGwoc29ja2V0cywgMSwgaGFzX3RpbWVvdXQ/bXNf dG9fbmV4dDogLTEpOw0KKyAgICB9DQogI2Vsc2UNCi0gICAgICAgICAgICBm ZF9zZXQgcmZkOw0KLSAgICAgICAgICAgIGZkX3NldCB3ZmQ7DQotICAgICAg ICAgICAgZmRfc2V0ICp3cml0ZWZkID0gTlVMTDsNCi0gICAgICAgICAgICBm ZF9zZXQgKnJlYWRmZCA9IE5VTEw7DQotICAgICAgICAgICAgc3RydWN0IHRp bWV2YWwgdHY7DQotDQotICAgICAgICAgICAgaWYgKHNlc3Npb24tPmFwaV90 aW1lb3V0ID4gMCAmJg0KLSAgICAgICAgICAgICAgICAoc2Vjb25kc190b19u ZXh0ID09IDAgfHwNCi0gICAgICAgICAgICAgICAgIHNlY29uZHNfdG9fbmV4 dCA+IHNlc3Npb24tPmFwaV90aW1lb3V0KSkgew0KLSAgICAgICAgICAgICAg ICB0aW1lX3Qgbm93ID0gdGltZSAoTlVMTCk7DQotICAgICAgICAgICAgICAg IGxvbmcgZWxhcHNlZF9tcyA9IChsb25nKSgxMDAwKmRpZmZ0aW1lKHN0YXJ0 X3RpbWUsIG5vdykpOw0KLSAgICAgICAgICAgICAgICBpZiAoZWxhcHNlZF9t cyA+IHNlc3Npb24tPmFwaV90aW1lb3V0KSB7DQotICAgICAgICAgICAgICAg ICAgICBzZXNzaW9uLT5lcnJfY29kZSA9IExJQlNTSDJfRVJST1JfVElNRU9V VDsNCi0gICAgICAgICAgICAgICAgICAgIHJldHVybiBMSUJTU0gyX0VSUk9S X1RJTUVPVVQ7DQotICAgICAgICAgICAgICAgIH0NCi0gICAgICAgICAgICAg ICAgdHYudHZfc2VjID0gKHNlc3Npb24tPmFwaV90aW1lb3V0IC0gZWxhcHNl ZF9tcykgLyAxMDAwOw0KLSAgICAgICAgICAgICAgICB0di50dl91c2VjID0g KChzZXNzaW9uLT5hcGlfdGltZW91dCAtIGVsYXBzZWRfbXMpICUgMTAwMCkg Kg0KLSAgICAgICAgICAgICAgICAgICAgMTAwMDsNCi0gICAgICAgICAgICAg ICAgaGFzX3RpbWVvdXQgPSAxOw0KLSAgICAgICAgICAgIH0NCi0gICAgICAg ICAgICBlbHNlIGlmIChzZWNvbmRzX3RvX25leHQgPiAwKSB7DQotICAgICAg ICAgICAgICAgIHR2LnR2X3NlYyA9IHNlY29uZHNfdG9fbmV4dDsNCi0gICAg ICAgICAgICAgICAgdHYudHZfdXNlYyA9IDA7DQotICAgICAgICAgICAgICAg IGhhc190aW1lb3V0ID0gMTsNCi0gICAgICAgICAgICB9DQotICAgICAgICAg ICAgZWxzZQ0KLSAgICAgICAgICAgICAgICBoYXNfdGltZW91dCA9IDA7DQot DQotICAgICAgICAgICAgaWYoZGlyICYgTElCU1NIMl9TRVNTSU9OX0JMT0NL X0lOQk9VTkQpIHsNCi0gICAgICAgICAgICAgICAgRkRfWkVSTygmcmZkKTsN Ci0gICAgICAgICAgICAgICAgRkRfU0VUKHNlc3Npb24tPnNvY2tldF9mZCwg JnJmZCk7DQotICAgICAgICAgICAgICAgIHJlYWRmZCA9ICZyZmQ7DQotICAg ICAgICAgICAgfQ0KLQ0KLSAgICAgICAgICAgIGlmKGRpciAmIExJQlNTSDJf U0VTU0lPTl9CTE9DS19PVVRCT1VORCkgew0KLSAgICAgICAgICAgICAgICBG RF9aRVJPKCZ3ZmQpOw0KLSAgICAgICAgICAgICAgICBGRF9TRVQoc2Vzc2lv bi0+c29ja2V0X2ZkLCAmd2ZkKTsNCi0gICAgICAgICAgICAgICAgd3JpdGVm ZCA9ICZ3ZmQ7DQotICAgICAgICAgICAgfQ0KKyAgICB7DQorICAgICAgICBm ZF9zZXQgcmZkOw0KKyAgICAgICAgZmRfc2V0IHdmZDsNCisgICAgICAgIGZk X3NldCAqd3JpdGVmZCA9IE5VTEw7DQorICAgICAgICBmZF9zZXQgKnJlYWRm ZCA9IE5VTEw7DQorICAgICAgICBzdHJ1Y3QgdGltZXZhbCB0djsNCisNCisg ICAgICAgIHR2LnR2X3NlYyA9IG1zX3RvX25leHQgLyAxMDAwOw0KKyAgICAg ICAgdHYudHZfdXNlYyA9IChtc190b19uZXh0IC0gdHYudHZfc2VjKjEwMDAp ICogMTAwMDsNCisNCisgICAgICAgIGlmKGRpciAmIExJQlNTSDJfU0VTU0lP Tl9CTE9DS19JTkJPVU5EKSB7DQorICAgICAgICAgICAgRkRfWkVSTygmcmZk KTsNCisgICAgICAgICAgICBGRF9TRVQoc2Vzc2lvbi0+c29ja2V0X2ZkLCAm cmZkKTsNCisgICAgICAgICAgICByZWFkZmQgPSAmcmZkOw0KKyAgICAgICAg fQ0KIA0KLSAgICAgICAgICAgIHJjID0gc2VsZWN0KHNlc3Npb24tPnNvY2tl dF9mZCArIDEsIHJlYWRmZCwgd3JpdGVmZCwgTlVMTCwNCi0gICAgICAgICAg ICAgICAgICAgICAgICBoYXNfdGltZW91dCA/ICZ0diA6IE5VTEwpOw0KLSNl bmRpZg0KKyAgICAgICAgaWYoZGlyICYgTElCU1NIMl9TRVNTSU9OX0JMT0NL X09VVEJPVU5EKSB7DQorICAgICAgICAgICAgRkRfWkVSTygmd2ZkKTsNCisg ICAgICAgICAgICBGRF9TRVQoc2Vzc2lvbi0+c29ja2V0X2ZkLCAmd2ZkKTsN CisgICAgICAgICAgICB3cml0ZWZkID0gJndmZDsNCiAgICAgICAgIH0NCi0g ICAgfQ0KIA0KKyAgICAgICAgcmMgPSBzZWxlY3Qoc2Vzc2lvbi0+c29ja2V0 X2ZkICsgMSwgcmVhZGZkLCB3cml0ZWZkLCBOVUxMLA0KKyAgICAgICAgICAg ICAgICAgICAgaGFzX3RpbWVvdXQgPyAmdHYgOiBOVUxMKTsNCisgICAgfQ0K KyNlbmRpZg0KICAgICBpZihyYyA8PSAwKSB7DQogICAgICAgICAvKiB0aW1l b3V0IChvciBlcnJvciksIGJhaWwgb3V0IHdpdGggYSB0aW1lb3V0IGVycm9y ICovDQogICAgICAgICBzZXNzaW9uLT5lcnJfY29kZSA9IExJQlNTSDJfRVJS T1JfVElNRU9VVDsNCi0tIA0KMS43LjUuMw0KDQo= --1129329158-2020387728-1306878700=:26627 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel --1129329158-2020387728-1306878700=:26627-- From libssh2-devel-bounces@cool.haxx.se Tue May 31 23:52:51 2011 Return-Path: Received: from www.haxx.se (localhost.localdomain [127.0.0.1]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLqnk4012509; Tue, 31 May 2011 23:52:51 +0200 Received: from giant.haxx.se (giant.haxx.se [80.67.6.50]) by giant.haxx.se (8.14.4/8.14.4/Debian-2) with ESMTP id p4VLqmWp012490 for ; Tue, 31 May 2011 23:52:48 +0200 Date: Tue, 31 May 2011 23:52:48 +0200 (CEST) From: Daniel Stenberg X-X-Sender: dast@giant.haxx.se To: libssh2 development Subject: Re: Core dump when authentication fails? In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) X-fromdanielhimself: yes MIME-Version: 1.0 X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.3.8 (giant.haxx.se [127.0.0.1]); Tue, 31 May 2011 23:52:51 +0200 (CEST) X-Greylist: Default is to whitelist mail, not delayed by milter-greylist-4.3.8 (giant.haxx.se [80.67.6.50]); Tue, 31 May 2011 23:52:48 +0200 (CEST) X-BeenThere: libssh2-devel@cool.haxx.se X-Mailman-Version: 2.1.14 Precedence: list Reply-To: libssh2 development List-Id: libssh2 development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: libssh2-devel-bounces@cool.haxx.se Sender: libssh2-devel-bounces@cool.haxx.se On Tue, 31 May 2011, Saqib Ali wrote: > I'll still test your changes. But can I get them from the daily snapshot for > June 1, 2011? Sure! I've pushed them so they'll be included in upcoming snapshots. -- / daniel.haxx.se _______________________________________________ libssh2-devel http://cool.haxx.se/cgi-bin/mailman/listinfo/libssh2-devel