public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
* Re: Suppressing code motion of constant values...
@ 1998-04-24 11:56 Joern Rennecke
  0 siblings, 0 replies; 4+ messages in thread
From: Joern Rennecke @ 1998-04-24 11:56 UTC (permalink / raw)
  To: lehotsky; +Cc: egcs

> There was a code snippet and discussion of the idea of
> suppressing hoisting constants outside of loops at
> lower optimization levels.
> 
> I thought at the time that this wasn't a winner for me, but in
> looking at some generated code for the SHARC DSP chip, it
> appears that this is causing me to use more registers than
> I need.
> 
> Can someone resend that code fragment?  It was a patch to loop.c
> if I remember correctly.

----- Forwarded message from Joern Rennecke -----

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Suppressing code motion of constant values...
  1998-04-24  6:54 Alan Lehotsky
  1998-04-24 11:56 ` Jeffrey A Law
@ 1998-04-24 16:23 ` Jim Wilson
  1 sibling, 0 replies; 4+ messages in thread
From: Jim Wilson @ 1998-04-24 16:23 UTC (permalink / raw)
  To: Alan Lehotsky; +Cc: egcs

It was a suggestion for the -Os (optimize for space) option, not one of
the speed optimization options.

The message was posted 03/30 by Joern Rennecke, with subject line
"New size optimization".

Jim

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Suppressing code motion of constant values...
  1998-04-24  6:54 Alan Lehotsky
@ 1998-04-24 11:56 ` Jeffrey A Law
  1998-04-24 16:23 ` Jim Wilson
  1 sibling, 0 replies; 4+ messages in thread
From: Jeffrey A Law @ 1998-04-24 11:56 UTC (permalink / raw)
  To: Alan Lehotsky; +Cc: egcs

  In message < l03130300b1665719a9bd@[209.61.104.176] >you write:
  > 
  > There was a code snippet and discussion of the idea of
  > suppressing hoisting constants outside of loops at
  > lower optimization levels.
  > 
  > I thought at the time that this wasn't a winner for me, but in
  > looking at some generated code for the SHARC DSP chip, it
  > appears that this is causing me to use more registers than
  > I need.
  > 
  > Can someone resend that code fragment?  It was a patch to loop.c
  > if I remember correctly.
It was a patch to loop and you should be able to find it in the
archives.  Probably the best way would be to use the by

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Suppressing code motion of constant values...
@ 1998-04-24  6:54 Alan Lehotsky
  1998-04-24 11:56 ` Jeffrey A Law
  1998-04-24 16:23 ` Jim Wilson
  0 siblings, 2 replies; 4+ messages in thread
From: Alan Lehotsky @ 1998-04-24  6:54 UTC (permalink / raw)
  To: egcs

There was a code snippet and discussion of the idea of
suppressing hoisting constants outside of loops at
lower optimization levels.

I thought at the time that this wasn't a winner for me, but in
looking at some generated code for the SHARC DSP chip, it
appears that this is causing me to use more registers than
I need.

Can someone resend that code fragment?  It was a patch to loop.c
if I remember correctly.

-- Al Lehotsky



------------------------------------------------------------------------

		    Quality Software Management
		http://www.tiac.net/users/lehotsky
			lehotsky@tiac.net
			(978)371-2219 Voice
			(978)371-2382 Fax/Data

	Software Process Improvement and Management Consulting
	     Language Design and Compiler Implementation



^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~1998-04-24 16:23 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-04-24 11:56 Suppressing code motion of constant values Joern Rennecke
  -- strict thread matches above, loose matches on Subject: below --
1998-04-24  6:54 Alan Lehotsky
1998-04-24 11:56 ` Jeffrey A Law
1998-04-24 16:23 ` Jim Wilson

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).