RE: [Libvir] Libvirt on windows

Hi, How stable is this compiled DLL? I understand that this is from version 0.4.0, Do you have a 0.3 version that is not "experimental" (as you put it ...) By the way I'm using the DLL without problems if anyone needs the dynamic binding code, I can send it to him in order to save time. Gaby

[Libvir] Re: libvirt on mingw

Brecht Sanders wrote: Hello again, Note that I am using gcc (mingw) 3.4.5, and not 4 as you recommend. The attached patch got rid of the warnings though. Maybe there is a compiler switch to fix it in a cleaner way though. Anyway, with the attached patch it does compile. Basically it casts to void

Re: [Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

On Wed, Jan 30, 2008 at 07:00:30PM +0100, Jim Meyering wrote: > "Daniel P. Berrange" <[EMAIL PROTECTED]> wrote: > > > > Not, its a bug in virUnrefDomain/Network - it calls mutex_lock() twice > > in one codepath, instead of calling unlock(). > > > > Of course your patch to avoid the memory leak is

Re: [Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

"Daniel P. Berrange" <[EMAIL PROTECTED]> wrote: > On Wed, Jan 30, 2008 at 06:36:37PM +0100, Jim Meyering wrote: >> "Daniel P. Berrange" <[EMAIL PROTECTED]> wrote: >> >> > On Wed, Jan 30, 2008 at 06:25:19PM +0100, Jim Meyering wrote: >> >> Without the following patch, this command would hang >> >>

Re: [Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

On Wed, Jan 30, 2008 at 06:36:37PM +0100, Jim Meyering wrote: > "Daniel P. Berrange" <[EMAIL PROTECTED]> wrote: > > > On Wed, Jan 30, 2008 at 06:25:19PM +0100, Jim Meyering wrote: > >> Without the following patch, this command would hang > >> > >> printf 'domuuid fc4\ndomstate fc4\n' \ > >>

Re: [Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

"Daniel P. Berrange" <[EMAIL PROTECTED]> wrote: > On Wed, Jan 30, 2008 at 06:25:19PM +0100, Jim Meyering wrote: >> Without the following patch, this command would hang >> >> printf 'domuuid fc4\ndomstate fc4\n' \ >> | ./virsh --connect test://$PWD/../docs/testnode.xml >> >> with this sta

Re: [Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

On Wed, Jan 30, 2008 at 06:25:19PM +0100, Jim Meyering wrote: > Without the following patch, this command would hang > > printf 'domuuid fc4\ndomstate fc4\n' \ > | ./virsh --connect test://$PWD/../docs/testnode.xml > > with this stack trace: > > __lll_lock_wait ... > _L_lock_10

[Libvir] [PATCH] avoid virsh hang due to missing virDomainFree(dom) call

Without the following patch, this command would hang printf 'domuuid fc4\ndomstate fc4\n' \ | ./virsh --connect test://$PWD/../docs/testnode.xml with this stack trace: __lll_lock_wait ... _L_lock_105 ... __pthread_mutex_lock ... virUnrefDomain (domain=0x6a8b30) at hash.

[Libvir] Re: libvirt on mingw

Brecht Sanders wrote: Hi, I saw on the following link: http://www.mail-archive.com/libvir-list@redhat.com/msg04103.html that you are also trying to compile libvirt on win32. I'm also attempting to do this, and I guess I got stuck at the same point your post was about. Have you in the mean tim

Re: [Libvir] [PATCH] Fix USB device name mis-conversion from S-Expr to XML

On Wed, Jan 30, 2008 at 11:22:16AM +0900, Hiroyuki Kaguchi wrote: > Would you give me a comment on this patch? If not, please apply it. > > > "tablet" and "mouse" are set as a value of the > > /local/domain//image/hvm/usbdevice, > > but libvirt expects "usbdevice" and "usbmouse" as a value. > > Th

Re: [Libvir] virsh capabilities cmd: plug a leak

On Wed, Jan 30, 2008 at 04:39:37PM +0100, Jim Meyering wrote: > Here's another: > > $ valgrind --leak-check=full \ > ./virsh --connect test://$PWD/../docs/testnode.xml capabilities > ... > ==16392== 368 bytes in 1 blocks are definitely lost in loss record 13 of 27 > ==16392==at 0x4

[Libvir] virsh capabilities cmd: plug a leak

Here's another: $ valgrind --leak-check=full \ ./virsh --connect test://$PWD/../docs/testnode.xml capabilities ... ==16392== 368 bytes in 1 blocks are definitely lost in loss record 13 of 27 ==16392==at 0x4A059F6: malloc (vg_replace_malloc.c:149) ==16392==by 0x387BC7CC11: str

Re: [Libvir] plug small leaks upon failure

On Wed, Jan 30, 2008 at 04:07:32PM +0100, Jim Meyering wrote: > with more than 20 domains, I get this: > > $ valgrind --leak-check=full ./virsh --connect > test://$PWD/../docs/testnode.xml list > ... > error: failed to connect to the hypervisor > ... > ==13756== 5 bytes in 1 blocks are

[Libvir] plug small leaks upon failure

with more than 20 domains, I get this: $ valgrind --leak-check=full ./virsh --connect test://$PWD/../docs/testnode.xml list ... error: failed to connect to the hypervisor ... ==13756== 5 bytes in 1 blocks are definitely lost in loss record 1 of 32 ==13756==at 0x4A059F6: malloc (vg

[Libvir] avoid used-uninitialized errors in test driver

Here's another distracting and easily avoided error: $ valgrind --leak-check=full ./virsh --connect test://$PWD/../docs/testnode.xml list ==13150== Conditional jump or move depends on uninitialised value(s) ==13150==at 0x4177DE: testLoadDomain (test.c:324) ==13150==by 0x417BA4: te

[Libvir] [PATCH] test.c: Avoid segfault upon malloc failure, and plug a leak.

Testing exposed a tiny leak: $ valgrind --leak-check=full ./virsh --connect \ test://$PWD/../docs/testnode.xml save fc4 /dev/null ... ==11077== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 4 from 1) ==11077== malloc/free: in use at exit: 24,677 bytes in 367 blocks. ==11077=

[Libvir] libvirt.c: avoid a double-free upon do_open failure

With a contrived example using more than 20 (the max permitted by the testing framework) domains, I got a double-free error: ==4821== Invalid free() / delete / delete[] ==4821==at 0x4A0560B: free (vg_replace_malloc.c:233) ==4821==by 0x4167F1: virReleaseConnect (hash.c:717) ==4821==

Re: [Libvir] [PATCH] typo

Hi, Rich Thank you for your suggestion. Now I found, I switched off test utils since compilation failed with --with-test on MinGW. Since testutils failes compilation. I attach the failure log Thanks Atsushi SAKAI "Richard W.M. Jones" <[EMAIL PROTECTED]> wrote: > Atsushi SAKAI wrote: > > Hi, >