On Sun, Jan 06, 2019 at 11:15:20PM -0800, Dan Williams wrote:
On Sun, Jan 6, 2019 at 8:17 PM Michael S. Tsirkin <mst@xxxxxxxxxx> wrote:Hmm it's way cheaper even though IIRC it's measureable.
On Mon, Jan 07, 2019 at 11:53:41AM +0800, Jason Wang wrote:It would be interesting to see how expensive re-doing the address
On 2019/1/7 äå11:28, Michael S. Tsirkin wrote:
On Mon, Jan 07, 2019 at 10:19:03AM +0800, Jason Wang wrote:
On 2019/1/3 äå4:47, Michael S. Tsirkin wrote:OK so would you say it's really unsafe versus safe accesses?
On Sat, Dec 29, 2018 at 08:46:51PM +0800, Jason Wang wrote:On machine without SMAP (Sandy Bridge):
This series tries to access virtqueue metadata through kernel virtualWill review, thanks!
address instead of copy_user() friends since they had too much
overheads like checks, spec barriers or even hardware feature
toggling.
One questions that comes to mind is whether it's all about bypassing
stac/clac. Could you please include a performance comparison with
nosmap?
Before: 4.8Mpps
After: 5.2Mpps
Or would you say it's just a better written code?
It's the effect of removing speculation barrier.
You mean __uaccess_begin_nospec introduced by
commit 304ec1b050310548db33063e567123fae8fd0301
?
So fundamentally we do access_ok checks when supplying
the memory table to the kernel thread, and we should
do the spec barrier there.
Then we can just create and use a variant of uaccess macros that does
not include the barrier?
Or, how about moving the barrier into access_ok?
This way repeated accesses with a single access_ok get a bit faster.
CC Dan Williams on this idea.
limit check is compared to the speculation barrier. I.e. just switch
vhost_get_user() to use get_user() rather than __get_user(). That will
sanitize the pointer in the speculative path without a barrier.
Jason, would you like to try?
Although frankly __get_user being slower than get_user feels very wrong.
Not yet sure what to do exactly but would you agree?
I recall we had a convert access_ok() discussion with this result here:Sorry let me try to clarify. IIUC speculating access_ok once
https://lkml.org/lkml/2018/1/17/929
is harmless. As Linus said the problem is with "_subsequent_
accesses that can then be used to perturb the cache".
Thus:
1. if (!access_ok)
2. return
3. get_user
4. if (!access_ok)
5. return
6. get_user
Your proposal that Linus nacked was to effectively add a barrier after
lines 2 and 5 (also using the array_index_nospec trick for speed),
right? Unfortunately that needs a big API change.
I am asking about adding barrier_nospec within access_ok.
Thus effectively before lines 1 and 4.
access_ok will be slower but after all the point of access_ok is
to then access the same memory multiple times.
So we should be making __get_user faster and access_ok slower ...
...but it sounds like you are proposing a smaller scope fixup for theMaybe we'll have to. Except I hope vhost won't end up being the
vhost use case? Something like barrier_nospec() in the success path
for all vhost access_ok() checks and then a get_user() variant that
disables the barrier.
only user otherwise it will be hard to maintain.