Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 5 Jan 2020 15:05:41 -0500
From: Rich Felker <>
Subject: Re: [PATCH] math: move x86_64 fabs, fabsf to C with inline asm

On Sun, Jan 05, 2020 at 07:36:39PM +0300, Alexander Monakov wrote:
> ---
> Questions:
> Why are there amd64-specific fabs implementations in the first place?
> (Only) because GCC generated poor code for the generic C version?

I think so. It generates:

0000000000000000 <fabs>:
   0:   66 48 0f 7e c0          movq   %xmm0,%rax
   5:   48 ba ff ff ff ff ff    movabs $0x7fffffffffffffff,%rdx
   c:   ff ff 7f
   f:   48 21 d0                and    %rdx,%rax
  12:   48 89 44 24 f8          mov    %rax,-0x8(%rsp)
  17:   f2 0f 10 44 24 f8       movsd  -0x8(%rsp),%xmm0
  1d:   c3                      retq

> Do annotations for mask manipulation in the patch help? Any way to make
> them less ambiguous?

I think so. I like how you did individual asm statements with
dependency relationship between them so compiler could even schedule
them if it likes. I wonder if you could just write 0x7fffffffffffffff
as an operand and have the compiler load it, though.


Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.