All of lore.kernel.org
 help / color / mirror / Atom feed
* More __get_user_asm_ll32 ...
@ 2007-02-13 13:32 Ralf Baechle
  2007-02-13 16:06 ` Atsushi Nemoto
  0 siblings, 1 reply; 3+ messages in thread
From: Ralf Baechle @ 2007-02-13 13:32 UTC (permalink / raw)
  To: Franck Bui-Huu; +Cc: linux-mips

__get_user_asm_ll32 contains a junk move %0, $0 left over from an older
version of the code.  Ignoring fixups code and data this instruction
inflates the normal path in __get_user() by 50% which may explain much of
the size and performance you have meassures.  It also always clears
the error return, iow. __get_user_asm_ll32(..., &<something 64-bit>)
would never have returned an error.  ARGH.

  Ralf

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: More __get_user_asm_ll32 ...
  2007-02-13 13:32 More __get_user_asm_ll32 Ralf Baechle
@ 2007-02-13 16:06 ` Atsushi Nemoto
  2007-02-13 16:11   ` Ralf Baechle
  0 siblings, 1 reply; 3+ messages in thread
From: Atsushi Nemoto @ 2007-02-13 16:06 UTC (permalink / raw)
  To: ralf; +Cc: vagabon.xyz, linux-mips

On Tue, 13 Feb 2007 13:32:51 +0000, Ralf Baechle <ralf@linux-mips.org> wrote:
> __get_user_asm_ll32 contains a junk move %0, $0 left over from an older
> version of the code.  Ignoring fixups code and data this instruction
> inflates the normal path in __get_user() by 50% which may explain much of
> the size and performance you have meassures.  It also always clears
> the error return, iow. __get_user_asm_ll32(..., &<something 64-bit>)
> would never have returned an error.  ARGH.

Why is it clears the error return?  The "3:" label is just after the
junk move.

---
Atsushi Nemoto

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: More __get_user_asm_ll32 ...
  2007-02-13 16:06 ` Atsushi Nemoto
@ 2007-02-13 16:11   ` Ralf Baechle
  0 siblings, 0 replies; 3+ messages in thread
From: Ralf Baechle @ 2007-02-13 16:11 UTC (permalink / raw)
  To: Atsushi Nemoto; +Cc: vagabon.xyz, linux-mips

On Wed, Feb 14, 2007 at 01:06:15AM +0900, Atsushi Nemoto wrote:

> Why is it clears the error return?  The "3:" label is just after the
> junk move.

You're right and the final commit message reflects that.  So this is really
only a code size (320 bytes) and performance fix.

  Ralf

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2007-02-13 16:11 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-02-13 13:32 More __get_user_asm_ll32 Ralf Baechle
2007-02-13 16:06 ` Atsushi Nemoto
2007-02-13 16:11   ` Ralf Baechle

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.