On 16 March 2018 at 23:57, Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> wrote: > On Sat, Mar 10, 2018 at 03:21:45PM +0000, Ard Biesheuvel wrote: >> As reported by Sebastian, the way the arm64 NEON crypto code currently >> keeps kernel mode NEON enabled across calls into skcipher_walk_xxx() is >> causing problems with RT builds, given that the skcipher walk API may >> allocate and free temporary buffers it uses to present the input and >> output arrays to the crypto algorithm in blocksize sized chunks (where >> blocksize is the natural blocksize of the crypto algorithm), and doing >> so with NEON enabled means we're alloc/free'ing memory with preemption >> disabled. >> >> This was deliberate: when this code was introduced, each kernel_neon_begin() >> and kernel_neon_end() call incurred a fixed penalty of storing resp. >> loading the contents of all NEON registers to/from memory, and so doing >> it less often had an obvious performance benefit. However, in the mean time, >> we have refactored the core kernel mode NEON code, and now kernel_neon_begin() >> only incurs this penalty the first time it is called after entering the kernel, >> and the NEON register restore is deferred until returning to userland. This >> means pulling those calls into the loops that iterate over the input/output >> of the crypto algorithm is not a big deal anymore (although there are some >> places in the code where we relied on the NEON registers retaining their >> values between calls) >> >> So let's clean this up for arm64: update the NEON based skcipher drivers to >> no longer keep the NEON enabled when calling into the skcipher walk API. >> >> As pointed out by Peter, this only solves part of the problem. So let's >> tackle it more thoroughly, and update the algorithms to test the NEED_RESCHED >> flag each time after processing a fixed chunk of input. >> >> Given that this issue was flagged by the RT people, I would appreciate it >> if they could confirm whether they are happy with this approach. >> >> Changes since v4: >> - rebase onto v4.16-rc3 >> - apply the same treatment to new SHA512, SHA-3 and SM3 code that landed >> in v4.16-rc1 > > Looks good to me. If more work is needed we can always do > incremental fixes. > > Patches 1-22 applied. Thanks. Thanks Herbert. Apologies if this wasn't clear, but there are some cross dependencies with the arm64 tree, which receives non-trivial modifications in patches 10 and 11, which are subsequently depended upon by patches 12 - 23. Without acks from them, we should really not be merging this code yet, especially because I noticed a rebase issue in patch #10 (my bad). Would you mind reverting 10 - 22? I will revisit this asap, and try to get acks for the arm64 patches. If that means waiting for the next cycle, so be it. Thanks, Ard.
