On Mon, Jan 7, 2019 at 6:11 AM Michael S. Tsirkin <mst@xxxxxxxxxx> wrote: > > 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: > > > > > > On Mon, Jan 07, 2019 at 11:53:41AM +0800, Jason Wang wrote: > > > > > > > > On 2019/1/7 上午11:28, Michael S. Tsirkin wrote: > > > > > On Mon, Jan 07, 2019 at 10:19:03AM +0800, Jason Wang wrote: > > > > > > On 2019/1/3 上午4:47, Michael S. Tsirkin wrote: > > > > > > > On Sat, Dec 29, 2018 at 08:46:51PM +0800, Jason Wang wrote: > > > > > > > > This series tries to access virtqueue metadata through kernel virtual > > > > > > > > address instead of copy_user() friends since they had too much > > > > > > > > overheads like checks, spec barriers or even hardware feature > > > > > > > > toggling. > > > > > > > Will review, thanks! > > > > > > > One questions that comes to mind is whether it's all about bypassing > > > > > > > stac/clac. Could you please include a performance comparison with > > > > > > > nosmap? > > > > > > > > > > > > > On machine without SMAP (Sandy Bridge): > > > > > > > > > > > > Before: 4.8Mpps > > > > > > > > > > > > After: 5.2Mpps > > > > > OK so would you say it's really unsafe versus safe accesses? > > > > > 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. > > > > It would be interesting to see how expensive re-doing the address > > 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. > > Hmm it's way cheaper even though IIRC it's measureable. > 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? Agree. __get_user() being faster than get_user() defeats the whole point of converting code paths to the access_ok() + __get_user() pattern. > > > > I recall we had a convert access_ok() discussion with this result here: > > > > https://lkml.org/lkml/2018/1/17/929 > > Sorry let me try to clarify. IIUC speculating access_ok once > 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. If the barrier is before lines 1 and 4 then it offers no real protection as far I can see. It will start speculating again on the user controlled pointer value after the barrier resolves. > So we should be making __get_user faster and access_ok slower ... I agree, but then the barrier always needs to be after the access_ok() check unconditionally called in the return path from access_ok(). At that point it's back to the implementation that Linus nak'd, or I'm missing some other detail. ...but maybe if it is limited to a new access_ok_nospec() then the concern is addressed? Then rename current __get_user() to __get_user_nospec() and make a new __get_user() that is back to being optimal.
