Hi, Dan Please omit follwoing code from Masayuki's patch. It reverses Mizushima's @@ -511,10 +512,10 @@ xend_post(virConnectPtr xend, const char } else if ((ret == 202) && (strstr(content, "failed") != NULL)) { virXendError(xend, VIR_ERR_POST_FAILED, content); ret = -1; - } else if (((ret >= 200) && (ret <= 202)) && (strstr(content, "xend.err") != NULL)) { - /* This is to catch case of things like 'virsh dump Domain-0 foo' - * which returns a success code, but the word 'xend.err' - * in body to indicate error :-( + } else if ((ret == 202) && (strstr(content, "Cannot") != NULL)) { + /* This is to catch case of 'virsh dump Domain-0 foo' + * which returns a success code, but the word 'Cannot' + * in body to indicate error */ virXendError(xend, VIR_ERR_POST_FAILED, content); ret = -1; Thanks Atsushi SAKAI "Daniel P. Berrange" <berrange@redhat.com> wrote:
On Fri, Mar 02, 2007 at 08:42:59PM +0900, Masayuki Sunou wrote:
Hi Dan
This looks like a bug in XenD that should be reported upstrem. If the hypercall is given an invalid value it should reject it and not screw up the whole host.
I agree. I will consider it as a back log.
If we add this against the virConnectPtr object, we should name it
virConnectGetVcpuMax()
For consistency with other VCPU method naming. I wonder though, if we should
I contribute the patch that corrects the following again. ?$B!& Correction of name of method and argument --> Isn't the name bad? ?$B!& Correction of position of method
This patch looks good me - unless anyone else on the list has objections I'll commit it to CVS later today. (I'll tweak the name of the internal xenHypervisorGetMaxVcpus method to add in 'Domain')
Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=|
-- Libvir-list mailing list Libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list