On 24/07/15 12:41, Ian Jackson wrote:
> Andrew Cooper writes ("[PATCH] tools/libxl: Fixes to stream v2 task joining
> logic"):
>> During review of the libxl migration v2 series, I changes the task
>> joining logic, but clearly didn't think the result through
>> properly. This patch fixes several e
Andrew Cooper writes ("[PATCH] tools/libxl: Fixes to stream v2 task joining
logic"):
> During review of the libxl migration v2 series, I changes the task
> joining logic, but clearly didn't think the result through
> properly. This patch fixes several errors.
This would have been much easier to r
Andrew Cooper writes ("Re: [PATCH] tools/libxl: Fixes to stream v2 task joining
logic"):
> Migration v2 only ever has 1024 guest pages ever mapped at once (see
> xc_sr_save.c:write_batch() ), and doesn't leak the mapping, which means
> that the kernel decided that it was out of memory.
Maybe it
On 23/07/15 13:03, Wei Liu wrote:
> On Thu, Jul 23, 2015 at 12:42:25PM +0100, Andrew Cooper wrote:
>> On 23/07/15 12:38, Wei Liu wrote:
>>> On Thu, Jul 23, 2015 at 12:09:38PM +0100, Andrew Cooper wrote:
During review of the libxl migration v2 series, I changes the task
joining logic, but
On Thu, Jul 23, 2015 at 12:42:25PM +0100, Andrew Cooper wrote:
> On 23/07/15 12:38, Wei Liu wrote:
> > On Thu, Jul 23, 2015 at 12:09:38PM +0100, Andrew Cooper wrote:
> >> During review of the libxl migration v2 series, I changes the task
> >> joining logic, but clearly didn't think the result throu
On 23/07/15 12:38, Wei Liu wrote:
> On Thu, Jul 23, 2015 at 12:09:38PM +0100, Andrew Cooper wrote:
>> During review of the libxl migration v2 series, I changes the task
>> joining logic, but clearly didn't think the result through
>> properly. This patch fixes several errors.
>>
>> 1) Do not call c
On Thu, Jul 23, 2015 at 12:09:38PM +0100, Andrew Cooper wrote:
> During review of the libxl migration v2 series, I changes the task
> joining logic, but clearly didn't think the result through
> properly. This patch fixes several errors.
>
> 1) Do not call check_all_finished() in the success cases
During review of the libxl migration v2 series, I changes the task
joining logic, but clearly didn't think the result through
properly. This patch fixes several errors.
1) Do not call check_all_finished() in the success cases of
libxl__xc_domain_{save,restore}_done(). It serves no specific purpos
