mbox series

[for_v21,0/2] x86/sgx: Use SRCU and mmu_notifier

Message ID 20190711161625.15786-1-sean.j.christopherson@intel.com (mailing list archive)
Headers show
Series x86/sgx: Use SRCU and mmu_notifier | expand

Message

Sean Christopherson July 11, 2019, 4:16 p.m. UTC 
Add SRCU support in a standalone patch.  I included the mmu_notifier patch
as well so that you can compare approaches without having to wait on me
for additional input.

Sean Christopherson (2):
  x86/sgx: Use SRCU to protect mm_list during reclaim
  x86/sgx: Use mmu_notifier.release() instead of per-vma refcounting

 arch/x86/Kconfig                      |   2 +
 arch/x86/kernel/cpu/sgx/driver/main.c |  60 +++++----
 arch/x86/kernel/cpu/sgx/encl.c        | 182 +++++++++++++-------------
 arch/x86/kernel/cpu/sgx/encl.h        |  12 +-
 arch/x86/kernel/cpu/sgx/reclaim.c     |  71 ++++------
 5 files changed, 158 insertions(+), 169 deletions(-)

Comments

Jarkko Sakkinen July 11, 2019, 6:01 p.m. UTC | #1 
On Thu, Jul 11, 2019 at 09:16:23AM -0700, Sean Christopherson wrote:
> Add SRCU support in a standalone patch.  I included the mmu_notifier patch
> as well so that you can compare approaches without having to wait on me
> for additional input.

Might even take both now that I undertsand the vma_close() scenario.
Lets see...

/Jarkko
Jarkko Sakkinen July 11, 2019, 9:51 p.m. UTC | #2 
On Thu, Jul 11, 2019 at 09:16:23AM -0700, Sean Christopherson wrote:
> Add SRCU support in a standalone patch.  I included the mmu_notifier patch
> as well so that you can compare approaches without having to wait on me
> for additional input.

Can you rebase these to the latest master. I'll apply them once
I wake up (about 1AM now). Could just work around with raw patch
would need to get sleep first anyway. Thanks.

/Jarkko
Jarkko Sakkinen July 12, 2019, 4:17 a.m. UTC | #3 
On Fri, Jul 12, 2019 at 12:51:38AM +0300, Jarkko Sakkinen wrote:
> On Thu, Jul 11, 2019 at 09:16:23AM -0700, Sean Christopherson wrote:
> > Add SRCU support in a standalone patch.  I included the mmu_notifier patch
> > as well so that you can compare approaches without having to wait on me
> > for additional input.
> 
> Can you rebase these to the latest master. I'll apply them once
> I wake up (about 1AM now). Could just work around with raw patch
> would need to get sleep first anyway. Thanks.

No need, woke up early. I can do these manually with patch -p1.

Thnaks.

/Jarkko