[PATCH v3 0/2] Performance regression due to XSA-336

Boris Ostrovsky posted 2 patches 2 weeks, 1 day ago
Patches applied successfully (tree, apply log)
git fetch https://gitlab.com/xen-project/patchew/xen tags/patchew/1617052502-14181-1-git-send-email-boris.ostrovsky@oracle.com
Test gitlab-ci passed
xen/arch/x86/hvm/vpt.c        | 48 +++++++++++++++++++++++++++++++++----------
xen/include/asm-x86/hvm/vpt.h | 18 ++++++++++------
2 files changed, 49 insertions(+), 17 deletions(-)

[PATCH v3 0/2] Performance regression due to XSA-336

Posted by Boris Ostrovsky 2 weeks, 1 day ago
The first patch addresses performance regression introduced by XSA-336 fixes.
This patch could be considered as a candidate for inclusion in 4.15.

The second patch is a minor cleanup and can safely wait until after 4.15.


Boris Ostrovsky (2):
  x86/vpt: Do not take pt_migrate rwlock in some cases
  x86/vpt: Simplify locking argument to write_{un}lock

 xen/arch/x86/hvm/vpt.c        | 48 +++++++++++++++++++++++++++++++++----------
 xen/include/asm-x86/hvm/vpt.h | 18 ++++++++++------
 2 files changed, 49 insertions(+), 17 deletions(-)

-- 
1.8.3.1


Re: [PATCH v3 0/2] Performance regression due to XSA-336

Posted by Ian Jackson 2 weeks ago
Boris Ostrovsky writes ("[PATCH v3 0/2] Performance regression due to XSA-336"):
> The first patch addresses performance regression introduced by XSA-336 fixes.
> This patch could be considered as a candidate for inclusion in 4.15.

Thank you, but I think this is too late for 4.15.

Regards,
Ian.