- Devel - libvirt List Archives

[Libvir] Re: libvirt on mingw
by Richard W.M. Jones 04 Feb '08

04 Feb '08

Brecht Sanders wrote: > Did you have a win32 Python version of gconf? Nope. Please CC questions to libvir-list so that everyone can see questions & answers. Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903

1 0

[Libvir] virsh vncdisplay fix
by John Levon 03 Feb '08

03 Feb '08

I have an old HVM guest defined in 3.0.4, and now I'm running 3.1.2 and libvirt 0.4.0. The config looks like this: (domain (image (hvm (vnc 1) (vncdisplay 0) (device (vfb (location localhost:5900) (vncdisplay 0) (uuid ...) The libvirt code looks like this: 1757 /* Graphics device (HVM <= 3.0.4, or PV <= 3.0.3) vnc config */ 1758 if ((hvm && xendConfigVersion < 4) || 1759 (!hvm && xendConfigVersion < 3)) { 1760 tmp = sexpr_fmt_node(root, "domain/image/%s/vnc", hvm ? "hvm" : "linux"); So we don't see the 'vnc 1' at the top. But the code to parse device/vfb doesn't expect this config either: 1685 tmp = sexpr_node(node, "device/vfb/type"); The patch below replaces the later version check with something that copes correctly with the "upgrade" case. regards john 2008-01-15 John Levon <levon(a)movementarian.org> * src/xend_internal.c: fix VNC parsing of old Xen domains on newer Xen versions. diff -prauN libvirt-0.4.0/src/xend_internal.c libvirt-new/src/xend_internal.c --- libvirt-0.4.0/src/xend_internal.c 2007-12-17 15:05:27.000000000 -0800 +++ libvirt-new/src/xend_internal.c 2008-01-14 18:42:08.895060570 -0800 @@ -1371,6 +1371,7 @@ xend_parse_sexp_desc(virConnectPtr conn, unsigned char uuid[VIR_UUID_BUFLEN]; char uuidstr[VIR_UUID_STRING_BUFLEN]; int vif_index = 0; + int found_graphics = 0; if (root == NULL) { /* ERROR */ @@ -1683,6 +1684,7 @@ xend_parse_sexp_desc(virConnectPtr conn, if (tmp && !strcmp(tmp, "sdl")) { virBufferVSprintf(&buf, " <input type='mouse' bus='%s'/>\n", hvm ? "ps2": "xen"); virBufferAdd(&buf, " <graphics type='sdl'/>\n", 27); + found_graphics = 1; } else if (tmp && !strcmp(tmp, "vnc")) { int port = xenStoreDomainGetVNCPort(conn, domid); const char *listenAddr = sexpr_node(node, "device/vfb/vnclisten"); @@ -1700,6 +1702,7 @@ xend_parse_sexp_desc(virConnectPtr conn, if (keymap) virBufferVSprintf(&buf, " keymap='%s'", keymap); virBufferAdd(&buf, "/>\n", 3); + found_graphics = 1; } } } @@ -1751,8 +1754,7 @@ xend_parse_sexp_desc(virConnectPtr conn, } /* Graphics device (HVM <= 3.0.4, or PV <= 3.0.3) vnc config */ - if ((hvm && xendConfigVersion < 4) || - (!hvm && xendConfigVersion < 3)) { + if (!found_graphics) { tmp = sexpr_fmt_node(root, "domain/image/%s/vnc", hvm ? "hvm" : "linux"); if (tmp != NULL) { if (tmp[0] == '1') { @@ -1780,6 +1782,7 @@ xend_parse_sexp_desc(virConnectPtr conn, if (keymap) virBufferVSprintf(&buf, " keymap='%s'", keymap); virBufferAdd(&buf, "/>\n", 3); + found_graphics = 1; } } @@ -1789,6 +1792,7 @@ xend_parse_sexp_desc(virConnectPtr conn, if (tmp[0] == '1') { virBufferVSprintf(&buf, " <input type='mouse' bus='%s'/>\n", hvm ? "ps2" : "xen"); virBufferAdd(&buf, " <graphics type='sdl'/>\n", 27 ); + found_graphics = 1; } } }

2 5

[Libvir] [PATCH] Pull vc-list-files from gnulib, and update.
by Jim Meyering 01 Feb '08

01 Feb '08

Now that I've added the vc-list-files script to gnulib, this patch makes libvirt's bootstrap pull it from there. Changes from the previous version of the script: I reintroduced support for mercurial (only lightly tested), and rearranged the directory-existence tests so git is first. Here's the full patch, including a renaming in vasnprintf.c: ---------------- Pull vc-list-files from gnulib, and update. * bootstrap (gnulib_tool): Add the vc-list-files module. --- bootstrap | 3 ++- build-aux/vc-list-files | 19 +++++++++++++------ gnulib/lib/Makefile.am | 9 ++++++++- gnulib/lib/vasnprintf.c | 18 ++++++++++-------- gnulib/m4/gnulib-cache.m4 | 4 ++-- gnulib/m4/gnulib-comp.m4 | 1 + 6 files changed, 36 insertions(+), 18 deletions(-) diff --git a/bootstrap b/bootstrap index 751b244..1627086 100755 --- a/bootstrap +++ b/bootstrap @@ -77,7 +77,8 @@ $gnulib_tool \ --tests-base=gnulib/tests \ --import physmem getaddrinfo \ sys_stat vasprintf strndup \ - strsep poll gettext getpass + strsep poll gettext getpass \ + vc-list-files rm -f \ .gitignore \ diff --git a/build-aux/vc-list-files b/build-aux/vc-list-files index 72a6f54..1e158bb 100755 --- a/build-aux/vc-list-files +++ b/build-aux/vc-list-files @@ -1,7 +1,7 @@ #!/bin/sh -# List the specified version-controlled files. +# List version-controlled file names. -# Copyright (C) 2006, 2007 Free Software Foundation, Inc. +# Copyright (C) 2006-2008 Free Software Foundation, Inc. # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by @@ -18,7 +18,8 @@ # List the specified version-controlled files. -# With no argument, list them all. +# With no argument, list them all. With a single DIRECTORY argument, +# list the version-controlled files in that directory. # This script must be run solely from the top of a $srcdir build directory. # If there's an argument, it must be a single, "."-relative directory name. @@ -28,12 +29,17 @@ dir= case $# in 0) ;; 1) dir=$1 ;; - *) echo "$0: too many arguments" 1>&2; exit 1 ;; + *) echo "$0: too many arguments" 1>&2 + echo "Usage: $0 [DIR]" 1>&2; exit 1;; esac test "x$dir" = x && dir=. -if test -d CVS; then +if test -d .git; then + exec git ls-files "$dir" +elif test -d .hg; then + exec hg locate "$dir/*" +elif test -d CVS; then if test -x build-aux/cvsu; then build-aux/cvsu --find --types=AFGM "$dir" else @@ -46,5 +52,6 @@ if test -d CVS; then $(find ${*-*} -name Entries -print) /dev/null; fi else - git-ls-files "$dir" + echo "$0: Failed to determine type of version control used in "`pwd` 1>&2 + exit 1 fi diff --git a/gnulib/lib/Makefile.am b/gnulib/lib/Makefile.am index 64fc398..8f545ca 100644 --- a/gnulib/lib/Makefile.am +++ b/gnulib/lib/Makefile.am @@ -9,7 +9,7 @@ # the same distribution terms as the rest of that program. # # Generated by gnulib-tool. -# Reproduce by: gnulib-tool --import --dir=. --lib=libgnu --source-base=gnulib/lib --m4-base=gnulib/m4 --doc-base=doc --aux-dir=build-aux --with-tests --lgpl=2 --libtool --macro-prefix=gl getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf +# Reproduce by: gnulib-tool --import --dir=. --lib=libgnu --source-base=gnulib/lib --m4-base=gnulib/m4 --doc-base=doc --aux-dir=build-aux --with-tests --lgpl=2 --libtool --macro-prefix=gl getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf vc-list-files AUTOMAKE_OPTIONS = 1.5 gnits @@ -719,6 +719,13 @@ EXTRA_libgnu_la_SOURCES += asprintf.c vasprintf.c ## end gnulib module vasprintf +## begin gnulib module vc-list-files + + +EXTRA_DIST += $(top_srcdir)/build-aux/vc-list-files + +## end gnulib module vc-list-files + ## begin gnulib module wchar BUILT_SOURCES += $(WCHAR_H) diff --git a/gnulib/lib/vasnprintf.c b/gnulib/lib/vasnprintf.c index 6c8d5dc..205a9bf 100644 --- a/gnulib/lib/vasnprintf.c +++ b/gnulib/lib/vasnprintf.c @@ -1,5 +1,5 @@ /* vsprintf with automatic memory allocation. - Copyright (C) 1999, 2002-2007 Free Software Foundation, Inc. + Copyright (C) 1999, 2002-2008 Free Software Foundation, Inc. This program is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by @@ -95,7 +95,7 @@ #if (NEED_PRINTF_DOUBLE || NEED_PRINTF_INFINITE_DOUBLE) && !defined IN_LIBINTL # include <math.h> -# include "isnan.h" +# include "isnand.h" #endif #if (NEED_PRINTF_LONG_DOUBLE || NEED_PRINTF_INFINITE_LONG_DOUBLE) && !defined IN_LIBINTL @@ -106,7 +106,7 @@ #if (NEED_PRINTF_DIRECTIVE_A || NEED_PRINTF_DOUBLE) && !defined IN_LIBINTL # include <math.h> -# include "isnan.h" +# include "isnand.h" # include "printf-frexp.h" #endif @@ -236,7 +236,7 @@ decimal_point_char () static int is_infinite_or_zero (double x) { - return isnan (x) || x + x == x; + return isnand (x) || x + x == x; } #endif @@ -2327,7 +2327,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, # if NEED_PRINTF_DIRECTIVE_A || NEED_PRINTF_DOUBLE double arg = a.arg[dp->arg_index].a.a_double; - if (isnan (arg)) + if (isnand (arg)) { if (dp->conversion == 'A') { @@ -2676,7 +2676,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, if (dp->conversion == 'f' || dp->conversion == 'F') { double arg = a.arg[dp->arg_index].a.a_double; - if (!(isnan (arg) || arg + arg == arg)) + if (!(isnand (arg) || arg + arg == arg)) { /* arg is finite and nonzero. */ int exponent = floorlog10 (arg < 0 ? -arg : arg); @@ -3080,7 +3080,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, { double arg = a.arg[dp->arg_index].a.a_double; - if (isnan (arg)) + if (isnand (arg)) { if (dp->conversion >= 'A' && dp->conversion <= 'Z') { @@ -4327,7 +4327,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, if (prec_ourselves) { /* Handle the precision. */ - TCHAR_T *prec_ptr = + TCHAR_T *prec_ptr = # if USE_SNPRINTF (TCHAR_T *) (result + length); # else @@ -4654,6 +4654,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, not have this limitation. */ return result; +#if USE_SNPRINTF overflow: if (!(result == resultbuf || result == NULL)) free (result); @@ -4662,6 +4663,7 @@ VASNPRINTF (DCHAR_T *resultbuf, size_t *lengthp, CLEANUP (); errno = EOVERFLOW; return NULL; +#endif out_of_memory: if (!(result == resultbuf || result == NULL)) diff --git a/gnulib/m4/gnulib-cache.m4 b/gnulib/m4/gnulib-cache.m4 index d55e07a..357f170 100644 --- a/gnulib/m4/gnulib-cache.m4 +++ b/gnulib/m4/gnulib-cache.m4 @@ -15,11 +15,11 @@ # Specification in the form of a command-line invocation: -# gnulib-tool --import --dir=. --lib=libgnu --source-base=gnulib/lib --m4-base=gnulib/m4 --doc-base=doc --aux-dir=build-aux --with-tests --lgpl=2 --libtool --macro-prefix=gl getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf +# gnulib-tool --import --dir=. --lib=libgnu --source-base=gnulib/lib --m4-base=gnulib/m4 --doc-base=doc --aux-dir=build-aux --with-tests --lgpl=2 --libtool --macro-prefix=gl getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf vc-list-files # Specification in the form of a few gnulib-tool.m4 macro invocations: gl_LOCAL_DIR([]) -gl_MODULES([getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf]) +gl_MODULES([getaddrinfo getpass gettext physmem poll strndup strsep sys_stat vasprintf vc-list-files]) gl_AVOID([]) gl_SOURCE_BASE([gnulib/lib]) gl_M4_BASE([gnulib/m4]) diff --git a/gnulib/m4/gnulib-comp.m4 b/gnulib/m4/gnulib-comp.m4 index 1b91a1e..8896a38 100644 --- a/gnulib/m4/gnulib-comp.m4 +++ b/gnulib/m4/gnulib-comp.m4 @@ -225,6 +225,7 @@ AC_DEFUN([gltests_LIBSOURCES], [ AC_DEFUN([gl_FILE_LIST], [ build-aux/config.rpath build-aux/link-warning.h + build-aux/vc-list-files lib/alloca.in.h lib/asnprintf.c lib/asprintf.c -- 1.5.4.rc5.1.g0fa73

3 4

[Libvir] avoid used-uninitialized errors in test driver
by Jim Meyering 01 Feb '08

01 Feb '08

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: testLoadDomainFromFile (test.c:401) ==13150== by 0x418E52: testOpenFromFile (test.c:797) ==13150== by 0x4192C9: testOpen (test.c:913) ==13150== by 0x40FA20: do_open (libvirt.c:572) ==13150== by 0x40FDE4: virConnectOpenAuth (libvirt.c:681) ==13150== by 0x40D5B1: vshInit (virsh.c:4464) ==13150== by 0x40E67C: main (virsh.c:4985) ==13150== ==13150== Conditional jump or move depends on uninitialised value(s) ==13150== at 0x417F0D: testLoadNetwork (test.c:486) ==13150== by 0x4183AA: testLoadNetworkFromFile (test.c:577) ==13150== by 0x418F81: testOpenFromFile (test.c:822) ==13150== by 0x4192C9: testOpen (test.c:913) ==13150== by 0x40FA20: do_open (libvirt.c:572) ==13150== by 0x40FDE4: virConnectOpenAuth (libvirt.c:681) ==13150== by 0x40D5B1: vshInit (virsh.c:4464) ==13150== by 0x40E67C: main (virsh.c:4985) ==13150== ==13150== Conditional jump or move depends on uninitialised value(s) ==13150== at 0x41955B: testNumOfDomains (test.c:1020) ==13150== by 0x4104B9: virConnectNumOfDomains (libvirt.c:940) ==13150== by 0x405363: cmdList (virsh.c:557) ==13150== by 0x40C5F9: vshCommandRun (virsh.c:4032) ==13150== by 0x40E6AF: main (virsh.c:4991) Id Name State ---------------------------------- 1 fv0 running 2 fc4 running This patch fixes it: * src/test.c (testOpenFromFile): avoid used-uninitialized errors in test driver --- src/test.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/test.c b/src/test.c index 85170d9..fe5da41 100644 --- a/src/test.c +++ b/src/test.c @@ -672,7 +672,7 @@ static int testOpenFromFile(virConnectPtr conn, xmlNodePtr *domains, *networks = NULL; xmlXPathContextPtr ctxt = NULL; virNodeInfoPtr nodeInfo; - testConnPtr privconn = malloc(sizeof(*privconn)); + testConnPtr privconn = calloc(1, sizeof(*privconn)); if (!privconn) { testError(NULL, NULL, NULL, VIR_ERR_NO_MEMORY, "testConn"); return VIR_DRV_OPEN_ERROR; -- 1.5.4.rc5.1.g0fa73

2 1

[Libvir] libvirt.c: avoid a double-free upon do_open failure
by Jim Meyering 01 Feb '08

01 Feb '08

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== by 0x4168E7: virUnrefConnect (hash.c:746) ==4821== by 0x40FCEB: do_open (libvirt.c:621) ==4821== by 0x40FDF1: virConnectOpenAuth (libvirt.c:682) ==4821== by 0x40D5B1: vshInit (virsh.c:4464) ==4821== by 0x40E67C: main (virsh.c:4985) ==4821== Address 0x4C3BA68 is 0 bytes inside a block of size 60 free'd ==4821== at 0x4A0560B: free (vg_replace_malloc.c:233) ==4821== by 0x40FCB3: do_open (libvirt.c:618) ==4821== by 0x40FDF1: virConnectOpenAuth (libvirt.c:682) ==4821== by 0x40D5B1: vshInit (virsh.c:4464) ==4821== by 0x40E67C: main (virsh.c:4985) error: failed to connect to the hypervisor ... here's one way to fix it: diff --git a/src/libvirt.c b/src/libvirt.c index defadc1..c19565f 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -615,7 +615,6 @@ do_open (const char *name, return ret; failed: - free (ret->name); if (ret->driver) ret->driver->close (ret); if (uri) xmlFreeURI(uri); virUnrefConnect(ret); At first, rather than removing the offending free, I inserted this line just after it: ret->name = NULL; which avoids leaking ->name even if some driver-specific close function fails to clean up properly. But IMHO if such a function doesn't clean up properly then *it* should be fixed, not all callers.

2 1

[Libvir] [PATCH] test.c: Avoid segfault upon malloc failure, and plug a leak.
by Jim Meyering 01 Feb '08

01 Feb '08

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== malloc/free: 23,521 allocs, 23,154 frees, 3,424,316 bytes allocated. ==11077== For counts of detected errors, rerun with: -v ==11077== searching for pointers to 367 not-freed blocks. ==11077== checked 538,400 bytes. ==11077== ==11077== 4,000 bytes in 1 blocks are definitely lost in loss record 26 of 27 ==11077== at 0x4A059F6: malloc (vg_replace_malloc.c:149) ==11077== by 0x41C53A: virBufferNew (buf.c:132) ==11077== by 0x41AB22: testDomainDumpXML (test.c:1473) ==11077== by 0x41A1B5: testDomainSave (test.c:1284) ==11077== by 0x41120B: virDomainSave (libvirt.c:1342) ==11077== by 0x406169: cmdSave (virsh.c:1057) ==11077== by 0x40C5F9: vshCommandRun (virsh.c:4032) ==11077== by 0x40E6AF: main (virsh.c:4991) ==11077== ==11077== LEAK SUMMARY: ==11077== definitely lost: 4,000 bytes in 1 blocks. ==11077== possibly lost: 0 bytes in 0 blocks. ==11077== still reachable: 20,677 bytes in 366 blocks. ==11077== suppressed: 0 bytes in 0 blocks. Here's the fix: test.c: Avoid segfault upon malloc failure, and plug a leak. * src/test.c (testDomainSave): Time permitting, I will add tests to exercise as many of the commands as possible. --- src/test.c | 9 ++++++++- 1 files changed, 8 insertions(+), 1 deletions(-) diff --git a/src/test.c b/src/test.c index 35e41a3..85170d9 100644 --- a/src/test.c +++ b/src/test.c @@ -1,7 +1,7 @@ /* * test.c: A "mock" hypervisor for use by application unit tests * - * Copyright (C) 2006-2007 Red Hat, Inc. + * Copyright (C) 2006-2008 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or @@ -1281,6 +1281,11 @@ static int testDomainSave(virDomainPtr domain, GET_DOMAIN(domain, -1); xml = testDomainDumpXML(domain, 0); + if (xml == NULL) { + testError(domain->conn, domain, NULL, VIR_ERR_INTERNAL_ERROR, + "cannot allocate space for metadata"); + return (-1); + } if ((fd = open(path, O_CREAT|O_TRUNC|O_WRONLY, S_IRUSR|S_IWUSR)) < 0) { testError(domain->conn, domain, NULL, VIR_ERR_INTERNAL_ERROR, @@ -1303,9 +1308,11 @@ static int testDomainSave(virDomainPtr domain, if (write(fd, xml, len) != len) { testError(domain->conn, domain, NULL, VIR_ERR_INTERNAL_ERROR, "cannot write metadata"); + free(xml); close(fd); return (-1); } + free(xml); if (close(fd) < 0) { testError(domain->conn, domain, NULL, VIR_ERR_INTERNAL_ERROR, "cannot save domain data"); -- 1.5.4.rc5.1.g0fa73

2 1

[Libvir] [PATCH] fix compilation with USE_XEN == 0
by Guido Guenther 31 Jan '08

31 Jan '08

Hi, statstest.c needs to check if USE_XEN > 0, otherwise compilation breaks on --without-xen builds. Please apply if appropriate. Patch is against current cvs. Cheers, -- Guido

3 4

[Libvir] Re: libvirt on mingw
by Richard W.M. Jones 31 Jan '08

31 Jan '08

Brecht Sanders wrote: > I wasn't able to run virt-manager yet because of Python dependancy > issues, but I'm sure I'll be able to work that out. > Did you have virt-manager running on win32 yet? No, I wasn't able to get virt-manager running. Probably the same problems as you are having: building Python modules (specifically, libvirtmod.dll) which can work from the Win32 version of python, with gcc/MinGW, proves to be very hard. If you get it working, please let me know. Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903

1 0

[Libvir] virsh capabilities cmd: plug a leak
by Jim Meyering 31 Jan '08

31 Jan '08

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: strdup (strdup.c:43) ==16392== by 0x4194D2: testGetCapabilities (test.c:1006) ==16392== by 0x412B43: virConnectGetCapabilities (libvirt.c:2119) ==16392== by 0x408047: cmdCapabilities (virsh.c:2018) ==16392== by 0x40C5F9: vshCommandRun (virsh.c:4032) ==16392== by 0x40E6AF: main (virsh.c:4991) ==16392== Here's the patch: * src/virsh.c (cmdCapabilities): Plug a small leak. diff --git a/src/virsh.c b/src/virsh.c index c09dc8d..0d93fb6 100644 --- a/src/virsh.c +++ b/src/virsh.c @@ -2020,6 +2020,7 @@ cmdCapabilities (vshControl * ctl, vshCmd * cmd ATTRIBUTE_UNUSED) return FALSE; } vshPrint (ctl, "%s\n", caps); + free (caps); return TRUE; } -- 1.5.4.rc5.1.g0fa73

3 2

Re: [Libvir] Libvirt on windows
by Richard W.M. Jones 31 Jan '08

31 Jan '08

Gabriel Kaufmann wrote: > I tool libvirt-0.dll and I am using it in VC. > In order to do this, I had to dynamically load the DLL and map all its functions. This is somewhat outside my experience. My best suggestion (guess) is to try adding __declspec(dllexport) around some libvirt functions to see if this causes them to be exported to VC++-compiled programs. On gcc declspec doesn't appear to be necessary. See also: http://www.cygwin.com/ml/cygwin/2000-05/msg00430.html Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903

2 2