On Wed, Sep 21, 2011 at 10:55:37AM +0100, Daniel P. Berrange wrote:

On Tue, Sep 20, 2011 at 12:03:00PM +0800, Daniel Veillard wrote:

...

On Mon, Sep 19, 2011 at 08:15:18AM -0500, Adam Litke wrote:

...

On Mon, Sep 19, 2011 at 04:04:04PM +0800, Daniel Veillard wrote:

...

On Sun, Sep 18, 2011 at 09:37:22AM -0500, Adam Litke wrote: Hum, I wonder if remoteRelayDomainEventBlockJob shouldn't strdup the path string instead of using it directly in the remote_domain_event_block_job_msg block. As a result since we now free the datapointed by the xdr message within remoteDispatchDomainEventSend() , this errors wasn't shown before but leads to a double free now.

BTW it seems we don't check all allocations in the xdr code (on purpose ?) for example make_nonnull_domain() doesn't check a strdup.

Could you check the following patch ?

Yep, this seems to fix the problem (and an extra check with valgrind shows no memory leaks. Although I haven't verified it, the functions:

remoteRelayDomainEventIOError remoteRelayDomainEventIOErrorReason remoteRelayDomainEventGraphics

appear to have the same problem as well.

Right though they do far more allocations. I ended up pushing the following patch which tries to be a bit better on handling memory allocation error, but it's not fully complete yet:

Fix crash on events due to allocation errors

remoteRelayDomainEventBlockJob, remoteRelayDomainEventIOError, remoteRelayDomainEventIOErrorReason and remoteRelayDomainEventGraphics were using const string directly in rpc structure, before calling remoteDispatchDomainEventSend(). But that routine now frees up all the pointed allocated memory from the rpc structure and we end up with a double free. This now strdup() all the strings passed and provide mem_error goto labels to be used when an allocation error occurs. Note that the cleanup isn't completely finished because all relaying function also call make_nonnull_domain() which also allocate a string and never handle the error case. This patches doesn't try to address this as this is only error correctness a priori and touches far more functions in this module:

* daemon/remote.c: fix string allocations and memory error handling for remoteRelayDomainEventBlockJob, remoteRelayDomainEventIOError, remoteRelayDomainEventIOErrorReason and remoteRelayDomainEventGraphics

diff --git a/daemon/remote.c b/daemon/remote.c index 45244f8..74e759a 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -234,9 +234,13 @@ static int remoteRelayDomainEventIOError(virConnectPtr conn ATTRIBUTE_UNUSED,

/* build return data */ memset(&data, 0, sizeof data); + data.srcPath = strdup(srcPath); + if (data.srcPath == NULL) + goto mem_error; + data.devAlias = strdup(devAlias); + if (data.devAlias == NULL) + goto mem_error; make_nonnull_domain(&data.dom, dom); - data.srcPath = (char*)srcPath; - data.devAlias = (char*)devAlias; data.action = action;

remoteDispatchDomainEventSend(client, remoteProgram, @@ -244,6 +248,11 @@ static int remoteRelayDomainEventIOError(virConnectPtr conn ATTRIBUTE_UNUSED, (xdrproc_t)xdr_remote_domain_event_io_error_msg, &data);

return 0; +mem_error: + virReportOOMError(); + virFree(data.srcPath); + virFree(data.devAlias); + return -1;

s/virFree/VIR_FREE/ throughout all this code.

Yeah Eric pointed that out, I already commited his patch and the associated make syntax-check test :-) Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/