public inbox for ecos-discuss@sourceware.org
 help / color / mirror / Atom feed
* [ECOS] Mutex optimization for uncontested cases
@ 2001-09-24 18:40 Luoqi Chen
  2001-09-25 12:45 ` Jonathan Larmour
  0 siblings, 1 reply; 2+ messages in thread
From: Luoqi Chen @ 2001-09-24 18:40 UTC (permalink / raw)
  To: ecos-discuss

I have noticed very poor performance of stdio stream under eCos, about 4 times
slower than vxworks. Howerver I was able to improve the performance by a factor
of 2 by disabling the CYGSEM_LIBC_STDIO_THREAD_SAFE_STREAMS option. This points
to an inefficiency in the current mutex implementation that we do not optimize
for the most common cases of uncontested mutex acquisitions. I wonder if eCos
developers are aware of this problem and if there is any effort under way
trying to resolve it.

-lq

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [ECOS] Mutex optimization for uncontested cases
  2001-09-24 18:40 [ECOS] Mutex optimization for uncontested cases Luoqi Chen
@ 2001-09-25 12:45 ` Jonathan Larmour
  0 siblings, 0 replies; 2+ messages in thread
From: Jonathan Larmour @ 2001-09-25 12:45 UTC (permalink / raw)
  To: Luoqi Chen; +Cc: ecos-discuss

Luoqi Chen wrote:
> 
> I have noticed very poor performance of stdio stream under eCos, about 4 times
> slower than vxworks. Howerver I was able to improve the performance by a factor
> of 2 by disabling the CYGSEM_LIBC_STDIO_THREAD_SAFE_STREAMS option. This points
> to an inefficiency in the current mutex implementation that we do not optimize
> for the most common cases of uncontested mutex acquisitions. I wonder if eCos
> developers are aware of this problem and if there is any effort under way
> trying to resolve it.

That's not a priority for us right now. It's open source, so patches
welcome! Bear in mind that atomic test and set type operations aren't
portable :-|.

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2001-09-25 12:45 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-09-24 18:40 [ECOS] Mutex optimization for uncontested cases Luoqi Chen
2001-09-25 12:45 ` Jonathan Larmour

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).