Bug 13335 – cse of sub-expressions of zero_extend/sign_extend expressions

Bug 13335 - cse of sub-expressions of zero_extend/sign_extend expressions


Summary:	cse of sub-expressions of zero_extend/sign_extend expressions

Status:	RESOLVED DUPLICATE of bug 17104

Product:	gcc
Classification:	Unclassified
Component:	rtl-optimization
Version:	3.4.0

Importance:	P2 enhancement
Target Milestone:	---
Assigned To:	Fariborz Jahanian

URL:	http://gcc.gnu.org/ml/gcc-patches/200...
Keywords:	missed-optimization, patch

Depends on:
Blocks:
	Show dependency tree / graph

Reported:	2003-12-06 19:13 UTC by David Edelsohn
Modified:	2006-04-30 19:52 UTC (History)
CC List:	6 users (show)

See Also:
Host:
Target:
Build:
Known to work:
Known to fail:
Last reconfirmed:	2005-09-07 18:02:55

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description David Edelsohn 2003-12-06 19:13:38 UTC

http://gcc.gnu.org/ml/gcc-patches/2003-11/msg01611.html

Comment 1 Andrew Pinski 2003-12-06 19:54:12 UTC

Confirmed, if this patch is not accepted for 3.4, there is no need for it for
3.5 (for tree-ssa) as the 
testcase is optimized already at the tree level where you do not have the
problems of the register 
sizes.

Comment 2 Fariborz Jahanian 2003-12-06 19:58:21 UTC

tree-ssa fixes the small test case. But the large file in vortex still fails.
This patch is still needed 
and a revised versionb is being  reviewed at this time.

Comment 3 Andrew Pinski 2003-12-20 04:12:14 UTC

The patch is Fariborz's.

Comment 4 Dara Hazeghi 2004-01-17 08:14:29 UTC

Is the original patch posted for this still awaiting review, or did you have
plans to send an updated 
version?

Comment 5 David Edelsohn 2004-01-17 17:13:48 UTC

Subject: Re:  cse of sub-expressions of zero_extend/sign_extend expressions 

>>>>> dhazeghi at yahoo dot com writes:

Dara> Is the original patch posted for this still awaiting review, or did
Dara> you have plans to send an updated version?

	Yes, it still is awaiting review.

David

Comment 6 Andrew Pinski 2004-03-16 16:27:30 UTC

Pushing this one off to 3.4.1.

Comment 7 Mark Mitchell 2004-05-31 21:59:13 UTC

There's no evidence that this is a regression, so I've removed the target
indicator.

Comment 8 Steven Bosscher 2006-04-24 22:45:40 UTC

Has this one fallen through the cracks?

Comment 9 roger 2006-04-30 19:52:31 UTC

This bug is a duplicate of PR17104 which was fixed by Nathan Sidwell in
November 2004.  If you read comment #4, you'll notice that the failure of
CSE to handle the rs6000's rs6000_emit_move's zero_extends is identical.


*** This bug has been marked as a duplicate of 17104 ***