Re: [PATCH v7 3/3] x86: vdso: Wire up getrandom() vDSO implementation
From: "Jason A. Donenfeld" <Jason@zx2c4.com>
Date: 2022-11-28 00:19:53
Also in:
linux-crypto, linux-patches, lkml
From: "Jason A. Donenfeld" <Jason@zx2c4.com>
Date: 2022-11-28 00:19:53
Also in:
linux-crypto, linux-patches, lkml
On Sun, Nov 27, 2022 at 10:39:27PM +0000, Samuel Neves wrote:
On Sun, Nov 27, 2022 at 10:13 PM Jason A. Donenfeld [off-list ref] wrote:quoted
Hi Thomas, On Sat, Nov 26, 2022 at 12:08:41AM +0100, Thomas Gleixner wrote:quoted
Jason! On Thu, Nov 24 2022 at 17:55, Jason A. Donenfeld wrote:quoted
+++ b/arch/x86/entry/vdso/vgetrandom-chacha.S +/* + * Very basic SSE2 implementation of ChaCha20. Produces a given positive number + * of blocks of output with a nonce of 0, taking an input key and 8-byte + * counter. Importantly does not spill to the stack. Its arguments are:Basic or not.Heh, FYI I didn't mean "basic" here as in "doesn't need a review", but just that it's a straightforward technique and doesn't do any complicated multiblock pyrotechnics (which frankly aren't really needed).quoted
This needs a Reviewed-by from someone who understands SSE2 and ChaCha20 before this can go anywhere near the x86 tree.No problem. I'll see to it that somebody qualified gives this a review.I did look at this earlier. It looks fine. I would recommend changing + /* copy1,copy2 = key */ + movdqu 0x00(key),copy1 + movdqu 0x10(key),copy2 to + /* copy1,copy2 = key */ + movups 0x00(key),copy1 + movups 0x10(key),copy2 which has the same semantics, but saves a couple of code bytes. Likewise for + movdqu state0,0x00(output) + movdqu state1,0x10(output) + movdqu state2,0x20(output) + movdqu state3,0x30(output) Otherwise, Reviewed-by: Samuel Neves <redacted> # for vgetrandom-chacha.S
Thanks for the review and for the suggestion. Will do. Jason