On Tue, Mar 31, 2009 at 06:32:29AM -0400, Daniel Johnson wrote:
> It's now fixed on Mac OS X as well. Yay!
Same with Mandriva.
>>> Dan
--
http://www.MoveAnnouncer.com The web change of address service
Let webmasters know that your web site has moved
--
On Tue, 31 Mar 2009, Daniel Johnson wrote:
>> Yes. it fixes it for me. Now the 6XX tests pass with no problem.
>
> It's now fixed on Mac OS X as well. Yay!
Thanks both of you for all the testing.
I'm now back on track towards a libssh2 release, let's say in about 48 hours
to give some more time
On Mar 31, 2009, at 6:01 AM, Yang Tse wrote:
2009/3/31, Daniel Stenberg wrote:
Please update from CVS again and see if this might be what happened.
Yes. it fixes it for me. Now the 6XX tests pass with no problem.
I'll start a full rebuild shortly and submit the log to the */auto
page.
2009/3/31, Daniel Stenberg wrote:
> Please update from CVS again and see if this might be what happened.
Yes. it fixes it for me. Now the 6XX tests pass with no problem.
I'll start a full rebuild shortly and submit the log to the */auto page.
Cheers,
--
-=[Yang]=-
On Tue, 31 Mar 2009, Daniel Stenberg wrote:
> This would happen if the condition on src/channel.c:2450 never evaluates
> true, as then it'll simply skip the channel close phase.
Explaining things out loud is useful. After having written this it was clear
to me that the 'local.close' variable wa
On Tue, 31 Mar 2009, Yang Tse wrote:
> And after some convoluted memory allocation logging I can confirm that on
> this platform in all cases the leaked memory is the block that you've
> already pointed out previously which is allocated at sftp.c:621 for
> session->sftpInit_sftp.
>
> And except
2009/3/30, Daniel Stenberg wrote:
> Is there anyone else around that have curl built with libssh2 support who can
> see this or other problems when running the test suite in curl?
Ok, I also get the same memory leaks on cygwin.
TESTFAIL: These test cases failed: 600 602 604 608 609 610 611 612 6
On Mon, Mar 30, 2009 at 08:34:39PM +0200, Daniel Stenberg wrote:
> Is there anyone else around that have curl built with libssh2 support who can
> see this or other problems when running the test suite in curl?
I'm able to reproduce it on an i386 Mandriva system with GnuTLS:
http://curl.haxx.se/a
On Mar 30, 2009, at 2:34 PM, Daniel Stenberg wrote:
On Sat, 28 Mar 2009, Daniel Johnson wrote:
Thanks, I'll look into this once I have a computer again. It picked
a most
inconvienent time to die. Typical.
Is there anyone else around that have curl built with libssh2
support who can
see
On Sat, 28 Mar 2009, Daniel Johnson wrote:
> Thanks, I'll look into this once I have a computer again. It picked a most
> inconvienent time to die. Typical.
Is there anyone else around that have curl built with libssh2 support who can
see this or other problems when running the test suite in cu
On Sun, 29 Mar 2009, Peter Stuge wrote:
> Then simply match the return address with the Leak detected message to find
> out which alloc is unmatched. Maybe it is the same leak Jean-Louis found?
No unfortunately it isn't. That leak has existed for a long time and would've
triggered a failure in
On Mar 28, 2009, at 9:45 PM, Peter Stuge wrote:
> Daniel Johnson wrote:
** MEMORY FAILURE
Leak detected: memory still allocated: 188 bytes
At 30c264, there's 188 bytes.
allocated by ../../curl/lib/ssh.c:281
which is in libcurl's LIBSSH2_ALLOC_FUNC.
>
> ..
>> Too ba
Daniel Johnson wrote:
> >> ** MEMORY FAILURE
> >> Leak detected: memory still allocated: 188 bytes
> >> At 30c264, there's 188 bytes.
> >> allocated by ../../curl/lib/ssh.c:281
> >>
> >> which is in libcurl's LIBSSH2_ALLOC_FUNC.
..
> Too bad there's no OS X port of valgrind yet.
You can find out
On Mar 28, 2009, at 9:54 AM, Daniel Stenberg wrote:
> On Fri, 27 Mar 2009, Daniel Johnson wrote:
>
>> Well, I'm not getting segfaults, but I'm still getting test failures:
>>
>> test 600...[SFTP retrieval]
>>
>> ** MEMORY FAILURE
>> Leak detected: memory still allocated: 188 bytes
>> At 30c264,
On Fri, 27 Mar 2009, Daniel Johnson wrote:
> Well, I'm not getting segfaults, but I'm still getting test failures:
>
> test 600...[SFTP retrieval]
>
> ** MEMORY FAILURE
> Leak detected: memory still allocated: 188 bytes
> At 30c264, there's 188 bytes.
> allocated by ../../curl/lib/ssh.c:281
>
> wh
On Mar 27, 2009, at 6:26 PM, Daniel Stenberg wrote:
On Fri, 27 Mar 2009, Daniel Johnson wrote:
The tests work now, but there's a bigger problem. I tried running a
curl
test build with this and Bad Things happened. I got several more
segfaults
like this:
I fixed two three bugs that the c
On Fri, 27 Mar 2009, Daniel Johnson wrote:
> The tests work now, but there's a bigger problem. I tried running a curl
> test build with this and Bad Things happened. I got several more segfaults
> like this:
I fixed two three bugs that the curl tests showed and now I seem to be able to
run thr
On Fri, 27 Mar 2009, Daniel Johnson wrote:
> sh: line 1: 65407 Segmentation fault ../src/curl --output
> log/curl600.out --include --verbose --trace-time --key curl_client_key
> --pubkey curl_client_key.pub -u daniel:
> sftp://127.0.0.1:8999/build-8336/tests/log/file600.txt >> log/stdout60
On Mar 27, 2009, at 8:49 AM, Daniel Stenberg wrote:
On Fri, 27 Mar 2009, Daniel Johnson wrote:
I don't get the hang anymore, but I'm still getting segfaults. Both
tests
fail at the same place.
[...]
Maybe memory getting double-freed?
Thanks, it was actually using a struct pointer to a
On Fri, 27 Mar 2009, Daniel Johnson wrote:
> I don't get the hang anymore, but I'm still getting segfaults. Both tests
> fail at the same place.
[...]
> Maybe memory getting double-freed?
Thanks, it was actually using a struct pointer to a struct that had been freed
already... fixed now!
--
On Mar 27, 2009, at 5:39 AM, Daniel Stenberg wrote:
On Fri, 27 Mar 2009, Daniel Stenberg wrote:
Uh, I can and yes it's most probably related. I'll dig into it!
Thanks.
Fixed now, turned out I was just stupid as usual! ;-P
I don't get the hang anymore, but I'm still getting segfaults. Bot
Simon Josefsson writes:
> /Simon
>
> --
> ___
> libssh2-devel mailing list
> libssh2-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/libssh2-devel
Hm
On Fri, 27 Mar 2009, Simon Josefsson wrote:
> Anything holding up the v1.1 release now? People tend to test things more
> when there is a release...
Indeed. There's nothing in particular holding it up, so my plan is to release
it after the weekend if everything still looks fine then.
--
/
Daniel Stenberg writes:
> On Fri, 27 Mar 2009, Daniel Stenberg wrote:
>
>> Uh, I can and yes it's most probably related. I'll dig into it! Thanks.
>
> Fixed now, turned out I was just stupid as usual! ;-P
Works great now.
Anything holding up the v1.1 release now? People tend to test things
mor
Simon Josefsson writes:
> Daniel Stenberg writes:
>
>> My rather biggish overhaul of how to deal with the blocking API has now been
>> committed. Everyone please try out CVS HEAD and see how it runs for you!
>
> Running 'make check' just hangs for me, can you reproduce that and do
> you think i
On Fri, 27 Mar 2009, Daniel Stenberg wrote:
> Uh, I can and yes it's most probably related. I'll dig into it! Thanks.
Fixed now, turned out I was just stupid as usual! ;-P
--
/ daniel.haxx.se
--
On Fri, 27 Mar 2009, Simon Josefsson wrote:
> Running 'make check' just hangs for me, can you reproduce that and do you
> think it is related?
Uh, I can and yes it's most probably related. I'll dig into it! Thanks.
--
/ daniel.haxx.se
---
Daniel Stenberg writes:
> My rather biggish overhaul of how to deal with the blocking API has now been
> committed. Everyone please try out CVS HEAD and see how it runs for you!
Running 'make check' just hangs for me, can you reproduce that and do
you think it is related?
/Simon
-
Hey again friends,
My rather biggish overhaul of how to deal with the blocking API has now been
committed. Everyone please try out CVS HEAD and see how it runs for you!
Some words on my changes:
* The blocking mode is now only set for on API behavior. The actual socket
will unconditionall
29 matches
Mail list logo
