[CRIU] [RFC 0/2] mem: Don't mprotect VMA areas if there is no page to dump

Pavel Emelyanov xemul at parallels.com
Tue Feb 18 00:51:11 PST 2014


On 01/30/2014 04:47 PM, Cyrill Gorcunov wrote:
> Guys, this is what I'm ended with. Can't say I like how it looks now,
> so here is RFC stands. Please review, comments are welcome!
> 
> Cyrill Gorcunov (2):
>   mem: Make parasite_dump_pages_args to carry data at predefined place
>   mem: Don't mprotect VMA areas if there is no page to dump
> 
>  include/parasite.h | 18 +++++++++--
>  mem.c              | 89 ++++++++++++++++++++++++++++--------------------------
>  2 files changed, 63 insertions(+), 44 deletions(-)
> 

Guys, do we plan to rework this set? I think it's useful, I
see some non zero amount of empty VMAs in my experiments :)
This would help them a little.


More information about the CRIU mailing list