public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug optimization/14040] New: ARM cross compiler: error: could not split insn
@ 2004-02-06  5:40 phillim2 at comcast dot net
  2004-02-06  6:55 ` [Bug target/14040] [3.3 Regression] " pinskia at gcc dot gnu dot org
                   ` (19 more replies)
  0 siblings, 20 replies; 21+ messages in thread
From: phillim2 at comcast dot net @ 2004-02-06  5:40 UTC (permalink / raw)
  To: gcc-bugs

gcc spits out the following error: 

ni_morphology.c: In function `NI_BinaryErosion':
ni_morphology.c:277: error: could not split insn
(insn 3632 1279 3633 (nil) (cond_exec (ne (reg:CC 24 cc)
            (const_int 0 [0x0]))
        (set (reg:SI 2 r2)
            (plus:SI (reg/f:SI 13 sp)
                (const_int 3388 [0xd3c])))) 507 {ffssi2+1} (nil)
    (nil))
ni_morphology.c:277: internal compiler error: in final_scan_insn, at final.c:2793

gcc cmdline to reproduce: 

 /opt/cross/bin/arm-linux-gcc -fexpensive-optimizations  -fPIC ni_prep2.c -o
ni_prep2.o  -fif-conversion2  -fcse-follow-jumps -frerun-cse-after-loop
-fomit-frame-pointer

Preprocessed file to reproduce error:
http://www.linuxtr.net/gcc/ni_prep2.c
(Apologies for the size of the file, it took me long enough to reduce it this much)

Removal of any one of the above optimizations will allow the compile to complete. 

If -O2 is given the compiler will enter an infinite loop and eat memory until it
runs out. strace shows it looping on mmap(65536.....) and getting a new block of
memory for each mmap

Broken Versions: 3.3.2, 3.3.1. 
Known Working Versions: 2.95.3

Native i386 gcc works fine, so does x86_64 native.

-- 
           Summary: ARM cross compiler: error: could not split insn
           Product: gcc
           Version: 3.3.2
            Status: UNCONFIRMED
          Severity: critical
          Priority: P2
         Component: optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: phillim2 at comcast dot net
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
@ 2004-02-06  6:55 ` pinskia at gcc dot gnu dot org
  2004-02-07  3:02 ` phillim2 at comcast dot net
                   ` (18 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2004-02-06  6:55 UTC (permalink / raw)
  To: gcc-bugs



-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|optimization                |target
 GCC target triplet|                            |arm-elf
           Keywords|                            |ice-on-valid-code
      Known to fail|                            |3.3.2 3.3.1
            Summary|ARM cross compiler: error:  |[3.3 Regression] ARM cross
                   |could not split insn        |compiler: error: could not
                   |                            |split insn


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
  2004-02-06  6:55 ` [Bug target/14040] [3.3 Regression] " pinskia at gcc dot gnu dot org
@ 2004-02-07  3:02 ` phillim2 at comcast dot net
  2004-02-07 18:39 ` [Bug target/14040] [3.3/3.4 " pinskia at gcc dot gnu dot org
                   ` (17 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: phillim2 at comcast dot net @ 2004-02-07  3:02 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From phillim2 at comcast dot net  2004-02-07 03:02 -------
Built gcc-3.4 csl-arm-branch, glibc cvs, binutils cvs. Same result with
selective optimizations, error output: 

GNU C version 3.4.0 20031230 (CodeSourcery ARM Q4 2003) (arm-linux)
        compiled by GNU C version 3.3.1 (SuSE Linux).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
ni_morphology.c: In function `NI_BinaryErosion':
ni_morphology.c:277: error: could not split insn
(insn 3193 1132 3194 (cond_exec (ne (reg:CC cc)
            (const_int 0 [0x0]))
        (set (reg:SI r1)
            (plus:SI (reg/f:SI sp)
                (const_int 3380 [0xd34])))) 809 {ctzsi2+5} (nil)
    (nil))
ni_morphology.c:277: internal compiler error: in final_scan_insn, at final.c:2417

But, interestingly, adding -O2 does complete the compile and does not hit the
infinite loop problem. 


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
  2004-02-06  6:55 ` [Bug target/14040] [3.3 Regression] " pinskia at gcc dot gnu dot org
  2004-02-07  3:02 ` phillim2 at comcast dot net
@ 2004-02-07 18:39 ` pinskia at gcc dot gnu dot org
  2004-02-08 19:35 ` mmitchel at gcc dot gnu dot org
                   ` (16 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2004-02-07 18:39 UTC (permalink / raw)
  To: gcc-bugs



-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|3.3.2 3.3.1                 |3.3.2 3.3.1 3.4.0
      Known to work|                            |2.95.3
            Summary|[3.3 Regression] ARM cross  |[3.3/3.4 Regression] ARM
                   |compiler: error: could not  |cross compiler: error: could
                   |split insn                  |not split insn
   Target Milestone|---                         |3.3.3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (2 preceding siblings ...)
  2004-02-07 18:39 ` [Bug target/14040] [3.3/3.4 " pinskia at gcc dot gnu dot org
@ 2004-02-08 19:35 ` mmitchel at gcc dot gnu dot org
  2004-02-09 16:12 ` phillim2 at comcast dot net
                   ` (15 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-02-08 19:35 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-02-08 19:35 -------
I could not reproduce this problem with any of:

(1) CodeSourcery's ARM GNU Toolchain release.

(2) The 3.4 branch, configured for arm-linux or for arm-non-elf.

(3) The mainline, configure for arm-none-elf.  

(I didn't try arm-linux.)

In all cases, I was running on an i686-pc-linux-gnu host.

Therefore, I'm closing this as WORSKFORME.

Mike, if this still fails for you, would you please try to provide (a) the exact
 cc1 command-line used, which you can obtain by running "gcc" with "-v", and (b)
the host operating system you are using? 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |WORKSFORME


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (3 preceding siblings ...)
  2004-02-08 19:35 ` mmitchel at gcc dot gnu dot org
@ 2004-02-09 16:12 ` phillim2 at comcast dot net
  2004-02-09 17:46 ` rearnsha at gcc dot gnu dot org
                   ` (14 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: phillim2 at comcast dot net @ 2004-02-09 16:12 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From phillim2 at comcast dot net  2004-02-09 16:12 -------
Here's the different chains I've tested with:

Host 1: Athlon XP 2200, debian unstable, 2.6.0-test6 
Binary gcc dist, available from: http://www.openzaurus.org/official/toolchain/
2.95 from here works, 3.3.1 and 3.3.2 fail. 

3.3.2 info from gcc -v:

Thread model: posix
gcc version 3.3.2
 /opt/cross/lib/gcc-lib/arm-linux/3.3.2/cc1 -quiet -v -D__GNUC__=3 -
D__GNUC_MINOR__=3 -D__GNUC_PATCHLEVEL__=2 -D__ARM_ARCH_4T__ -D__PIC__ -D__pic__ 
ni_prep2.c -quiet -dumpbase ni_prep2.c -auxbase ni_prep2 -version -fexpensive-
optimizations -fPIC -fif-conversion2 -fcse-follow-jumps -frerun-cse-after-loop -
fomit-frame-pointer -o /tmp/ccbJMblz.s
GNU C version 3.3.2 (arm-linux)
        compiled by GNU C version 2.96 20000731 (Red Hat Linux 7.3 2.96-113).
GGC heuristics: --param ggc-min-expand=64 --param ggc-min-heapsize=64308

Host 2: Dual Opteron x86_64, Suse 9.0, 2.4.21-193 SMP kernel
Running the 32bit binaries noted above gives the same. 
However, I have a 3.3.2 x86_64->arm-linux tool chain built, gcc -v output: 

Reading specs from /usr/local/arm/lib/gcc-lib/arm-linux/3.3.2/specs
Configured with: ../gcc-3.3.2/configure --target=arm-linux --
prefix=/usr/local/arm --with-headers=../linux-2.6.0/include
Thread model: posix
gcc version 3.3.2
 /usr/local/arm/lib/gcc-lib/arm-linux/3.3.2/cc1 -quiet -v -D__GNUC__=3 -
D__GNUC_MINOR__=3 -D__GNUC_PATCHLEVEL__=2 -D__ARM_ARCH_4T__ -D__PIC__ -D__pic__ 
ni_prep2.c -quiet -dumpbase ni_prep2.c -auxbase ni_prep2 -version -fexpensive-
optimizations -fPIC -fif-conversion2 -fcse-follow-jumps -frerun-cse-after-loop -
fomit-frame-pointer -o /tmp/ccSTTXkS.s
GNU C version 3.3.2 (arm-linux)
        compiled by GNU C version 3.3.1 (SuSE Linux).
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072

The 3.4 codesourcery toolchain is x86_64->arm-linux too, gcc -v output: 

Reading specs from /opt/cross-3.4/lib/gcc/arm-linux/3.4.0/specs
Configured with: ./configure --prefix=/opt/cross-3.4 --host=x86_64 --target=arm-
linux --enable-languages=c
Thread model: posix
gcc version 3.4.0 20031230 (CodeSourcery ARM Q4 2003)
 /opt/cross-3.4/libexec/gcc/arm-linux/3.4.0/cc1 -quiet -v -D__ARM_ARCH_4T__ -
D__PIC__ -D__pic__ ni_prep2.c -quiet -dumpbase ni_prep2.c -auxbase ni_prep2 -
version -fexpensive-optimizations -fPIC -fif-conversion2 -fcse-follow-jumps -
frerun-cse-after-loop -fomit-frame-pointer -o /tmp/ccf7FJ1Y.s
ignoring duplicate directory "/opt/cross-3.4/lib/gcc/arm-
linux/3.4.0/../../../../arm-linux/include"
#include "..." search starts here:
#include <...> search starts here:
 /opt/cross-3.4/lib/gcc/arm-linux/3.4.0/include
 /opt/cross-3.4/lib/gcc/arm-linux/3.4.0/../../../../arm-linux/sys-include
End of search list.
GNU C version 3.4.0 20031230 (CodeSourcery ARM Q4 2003) (arm-linux)
        compiled by GNU C version 3.3.1 (SuSE Linux).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096

Mark: Do you have a binary tgz of a working i386 toolchain I could test with ?



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (4 preceding siblings ...)
  2004-02-09 16:12 ` phillim2 at comcast dot net
@ 2004-02-09 17:46 ` rearnsha at gcc dot gnu dot org
  2004-02-09 17:51 ` rearnsha at gcc dot gnu dot org
                   ` (13 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: rearnsha at gcc dot gnu dot org @ 2004-02-09 17:46 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From rearnsha at gcc dot gnu dot org  2004-02-09 17:46 -------
I can confirm the bug on a arm-elf cross with the following options:

cc1 -fexpensive-optimizations  -fPIC ni_prep2.c -fif-conversion2 
-fcse-follow-jumps -frerun-cse-after-loop -fomit-frame-pointer  -mno-apcs-frame

So reopened.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |UNCONFIRMED
         Resolution|WORKSFORME                  |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (5 preceding siblings ...)
  2004-02-09 17:46 ` rearnsha at gcc dot gnu dot org
@ 2004-02-09 17:51 ` rearnsha at gcc dot gnu dot org
  2004-02-15 12:46 ` gdr at gcc dot gnu dot org
                   ` (12 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: rearnsha at gcc dot gnu dot org @ 2004-02-09 17:51 UTC (permalink / raw)
  To: gcc-bugs



-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2004-02-09 17:51:16
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (6 preceding siblings ...)
  2004-02-09 17:51 ` rearnsha at gcc dot gnu dot org
@ 2004-02-15 12:46 ` gdr at gcc dot gnu dot org
  2004-02-22  1:40 ` rearnsha at gcc dot gnu dot org
                   ` (11 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: gdr at gcc dot gnu dot org @ 2004-02-15 12:46 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From gdr at gcc dot gnu dot org  2004-02-15 12:46 -------
Adjust milestone

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.3.3                       |3.3.4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (7 preceding siblings ...)
  2004-02-15 12:46 ` gdr at gcc dot gnu dot org
@ 2004-02-22  1:40 ` rearnsha at gcc dot gnu dot org
  2004-02-22 16:45 ` gdr at gcc dot gnu dot org
                   ` (10 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: rearnsha at gcc dot gnu dot org @ 2004-02-22  1:40 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From rearnsha at gcc dot gnu dot org  2004-02-22 01:40 -------
The problem here is that we have a define_insn_and_split that is predicable, but
splits are not generated for the auto-generated predicate variant.  The
constraints currently allow an immediate that is not valid for an add or sub
insn in the assembly file, the intention being that this can be split very late
on if necessary, but that means we fail if the patern has been made cond_exec.

There are several possible solutions to this, but I'm not sure yet which is the
best:

Most robust, but least desirable is to make the pattern non-predicable.
Second might be to remove the final alternative from the pattern, but that might
cause other failures if they rely on this pattern behaving this way.

Next would be to add an explicit split for this particular pattern when it is in
its cond_exec form

Finally, we might consider making the predicate-generation code auto-generate
suitable split patterns for the cond_exec cases it creates. (this is probably
the best work-around for 3.4 and, if we make it, 3.3.4)

I'm still evaluating the best solution here.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2004-02-09 17:51:16         |2004-02-22 01:40:25
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (8 preceding siblings ...)
  2004-02-22  1:40 ` rearnsha at gcc dot gnu dot org
@ 2004-02-22 16:45 ` gdr at gcc dot gnu dot org
  2004-03-01  1:57 ` mmitchel at gcc dot gnu dot org
                   ` (9 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: gdr at gcc dot gnu dot org @ 2004-02-22 16:45 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From gdr at gcc dot gnu dot org  2004-02-22 16:45 -------
(In reply to comment #7)

> I'm still evaluating the best solution here.

Richard --

  The branch is open again.  Did you find the way to go?

-- Gaby


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (9 preceding siblings ...)
  2004-02-22 16:45 ` gdr at gcc dot gnu dot org
@ 2004-03-01  1:57 ` mmitchel at gcc dot gnu dot org
  2004-03-22 21:49 ` mmitchel at gcc dot gnu dot org
                   ` (8 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-03-01  1:57 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-03-01 01:57 -------
Richard, I agree that modifying define_insn_and_split to generate splitters for
the predicated versions is the right fix.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rearnsha at arm dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (10 preceding siblings ...)
  2004-03-01  1:57 ` mmitchel at gcc dot gnu dot org
@ 2004-03-22 21:49 ` mmitchel at gcc dot gnu dot org
  2004-03-30 22:22 ` mmitchel at gcc dot gnu dot org
                   ` (7 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-03-22 21:49 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-03-22 21:49 -------
Working on a fix.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |mark at codesourcery dot com
                   |dot org                     |
             Status|NEW                         |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (11 preceding siblings ...)
  2004-03-22 21:49 ` mmitchel at gcc dot gnu dot org
@ 2004-03-30 22:22 ` mmitchel at gcc dot gnu dot org
  2004-03-31  9:55 ` rearnsha at arm dot com
                   ` (6 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-03-30 22:22 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-03-30 22:22 -------
I have a fix for this problem, but it is pretty invasive.  Therefore, I'm
retargeting this bug for 3.4.1.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.3.4                       |3.4.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (12 preceding siblings ...)
  2004-03-30 22:22 ` mmitchel at gcc dot gnu dot org
@ 2004-03-31  9:55 ` rearnsha at arm dot com
  2004-03-31 17:00 ` mark at codesourcery dot com
                   ` (5 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: rearnsha at arm dot com @ 2004-03-31  9:55 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From rearnsha at arm dot com  2004-03-31 09:55 -------
Subject: Re:  [3.3/3.4 Regression] ARM cross compiler:
	error: could not split insn

On Tue, 2004-03-30 at 23:22, mmitchel at gcc dot gnu dot org wrote:
> I have a fix for this problem, but it is pretty invasive.  Therefore, I'm
> retargeting this bug for 3.4.1.

Can the patch be put into the trunk for now, so that we can get some
shake-down?  It would be good if it could also be attached to the PR, so
that anyone experiencing this problem can try it out.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (13 preceding siblings ...)
  2004-03-31  9:55 ` rearnsha at arm dot com
@ 2004-03-31 17:00 ` mark at codesourcery dot com
  2004-04-01 23:30 ` cvs-commit at gcc dot gnu dot org
                   ` (4 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mark at codesourcery dot com @ 2004-03-31 17:00 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mark at codesourcery dot com  2004-03-31 17:00 -------
Subject: Re:  [3.3/3.4 Regression] ARM cross compiler: error:
 could not split insn

rearnsha at arm dot com wrote:

>------- Additional Comments From rearnsha at arm dot com  2004-03-31 09:55 -------
>Subject: Re:  [3.3/3.4 Regression] ARM cross compiler:
>	error: could not split insn
>
>On Tue, 2004-03-30 at 23:22, mmitchel at gcc dot gnu dot org wrote:
>  
>
>>I have a fix for this problem, but it is pretty invasive.  Therefore, I'm
>>retargeting this bug for 3.4.1.
>>    
>>
>
>Can the patch be put into the trunk for now, so that we can get some
>shake-down?  It would be good if it could also be attached to the PR, so
>that anyone experiencing this problem can try it out.
>  
>
Yes -- I just need to recover from jetlag, run some more detailed tests, 
create a ChangeLog, etc....  So, I hope to get to this soonish, but 
perhaps not instantly.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (14 preceding siblings ...)
  2004-03-31 17:00 ` mark at codesourcery dot com
@ 2004-04-01 23:30 ` cvs-commit at gcc dot gnu dot org
  2004-04-01 23:33 ` mmitchel at gcc dot gnu dot org
                   ` (3 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: cvs-commit at gcc dot gnu dot org @ 2004-04-01 23:30 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-04-01 23:30 -------
Subject: Bug 14040

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2004-04-01 23:30:27

Modified files:
	gcc            : ChangeLog genemit.c genrecog.c gensupport.c 
	gcc/config/arm : arm-protos.h arm.c arm.md 

Log message:
	PR target/14040
	* genemit.c (gen_split): Change prototype of generated code.
	* genrecog.c (write_action): Adjust prototype for and calls to
	gen_split_*.
	* gensupport.c (struct queue_elem): Add split field.
	(queue_pattern): Return a value.  Clear the split field.
	(process_rtx): Maintain an association between an insn and the
	split generated from it for a define_insn_and_split.
	(process_one_cond_exec): Generate a new split for a
	define_insn_and_split.
	* config/arm/arm-protos.h (arm_split_constant): Add insn
	parameter.
	(emit_constant_insn): New function.
	(arm_gen_constant): Use it.
	* config/arm/arm.md: Adjust calls to arm_split_constant.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.3330&r2=2.3331
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/genemit.c.diff?cvsroot=gcc&r1=1.91&r2=1.92
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/genrecog.c.diff?cvsroot=gcc&r1=1.139&r2=1.140
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/gensupport.c.diff?cvsroot=gcc&r1=1.47&r2=1.48
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm-protos.h.diff?cvsroot=gcc&r1=1.65&r2=1.66
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm.c.diff?cvsroot=gcc&r1=1.342&r2=1.343
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm.md.diff?cvsroot=gcc&r1=1.163&r2=1.164



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (15 preceding siblings ...)
  2004-04-01 23:30 ` cvs-commit at gcc dot gnu dot org
@ 2004-04-01 23:33 ` mmitchel at gcc dot gnu dot org
  2004-04-01 23:33 ` cvs-commit at gcc dot gnu dot org
                   ` (2 subsequent siblings)
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-04-01 23:33 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-04-01 23:33 -------
Fixed on the csl-arm-branch and on the mainline.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (16 preceding siblings ...)
  2004-04-01 23:33 ` mmitchel at gcc dot gnu dot org
@ 2004-04-01 23:33 ` cvs-commit at gcc dot gnu dot org
  2004-04-01 23:41 ` cvs-commit at gcc dot gnu dot org
  2004-06-12 22:51 ` mmitchel at gcc dot gnu dot org
  19 siblings, 0 replies; 21+ messages in thread
From: cvs-commit at gcc dot gnu dot org @ 2004-04-01 23:33 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-04-01 23:33 -------
Subject: Bug 14040

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	csl-arm-branch
Changes by:	mmitchel@gcc.gnu.org	2004-04-01 23:33:05

Modified files:
	gcc            : ChangeLog 

Log message:
	PR target/14040
	* genemit.c (gen_split): Change prototype of generated code.
	* genrecog.c (write_action): Adjust prototype for and calls to
	gen_split_*.
	* gensupport.c (struct queue_elem): Add split field.
	(queue_pattern): Return a value.  Clear the split field.
	(process_rtx): Maintain an association between an insn and the
	split generated from it for a define_insn_and_split.
	(process_one_cond_exec): Generate a new split for a
	define_insn_and_split.
	* config/arm/arm-protos.h (arm_split_constant): Add insn
	parameter.
	(emit_constant_insn): New function.
	(arm_gen_constant): Use it.
	* config/arm/arm.md: Adjust calls to arm_split_constant.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=2.1568.2.42&r2=2.1568.2.43



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (17 preceding siblings ...)
  2004-04-01 23:33 ` cvs-commit at gcc dot gnu dot org
@ 2004-04-01 23:41 ` cvs-commit at gcc dot gnu dot org
  2004-06-12 22:51 ` mmitchel at gcc dot gnu dot org
  19 siblings, 0 replies; 21+ messages in thread
From: cvs-commit at gcc dot gnu dot org @ 2004-04-01 23:41 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-04-01 23:41 -------
Subject: Bug 14040

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	csl-arm-branch
Changes by:	mmitchel@gcc.gnu.org	2004-04-01 23:41:32

Modified files:
	gcc            : genemit.c genrecog.c gensupport.c 
	gcc/config/arm : arm-protos.h arm.c arm.md 

Log message:
	PR target/14040
	* genemit.c (gen_split): Change prototype of generated code.
	* genrecog.c (write_action): Adjust prototype for and calls to
	gen_split_*.
	* gensupport.c (struct queue_elem): Add split field.
	(queue_pattern): Return a value.  Clear the split field.
	(process_rtx): Maintain an association between an insn and the
	split generated from it for a define_insn_and_split.
	(process_one_cond_exec): Generate a new split for a
	define_insn_and_split.
	* config/arm/arm-protos.h (arm_split_constant): Add insn
	parameter.
	(emit_constant_insn): New function.
	(arm_gen_constant): Use it.
	* config/arm/arm.md: Adjust calls to arm_split_constant.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/genemit.c.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.84.6.1&r2=1.84.6.2
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/genrecog.c.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.131.4.2&r2=1.131.4.3
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/gensupport.c.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.42.6.2&r2=1.42.6.3
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm-protos.h.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.60.4.8&r2=1.60.4.9
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm.c.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.303.2.21&r2=1.303.2.22
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/arm/arm.md.diff?cvsroot=gcc&only_with_tag=csl-arm-branch&r1=1.145.2.16&r2=1.145.2.17



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

* [Bug target/14040] [3.3/3.4 Regression] ARM cross compiler: error: could not split insn
  2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
                   ` (18 preceding siblings ...)
  2004-04-01 23:41 ` cvs-commit at gcc dot gnu dot org
@ 2004-06-12 22:51 ` mmitchel at gcc dot gnu dot org
  19 siblings, 0 replies; 21+ messages in thread
From: mmitchel at gcc dot gnu dot org @ 2004-06-12 22:51 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-06-12 22:51 -------
I think this is too invasive for 3.4.x, and it's already fixed in 3.5.0.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|3.4.1                       |3.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14040


^ permalink raw reply	[flat|nested] 21+ messages in thread

end of thread, other threads:[~2004-06-12 22:51 UTC | newest]

Thread overview: 21+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2004-02-06  5:40 [Bug optimization/14040] New: ARM cross compiler: error: could not split insn phillim2 at comcast dot net
2004-02-06  6:55 ` [Bug target/14040] [3.3 Regression] " pinskia at gcc dot gnu dot org
2004-02-07  3:02 ` phillim2 at comcast dot net
2004-02-07 18:39 ` [Bug target/14040] [3.3/3.4 " pinskia at gcc dot gnu dot org
2004-02-08 19:35 ` mmitchel at gcc dot gnu dot org
2004-02-09 16:12 ` phillim2 at comcast dot net
2004-02-09 17:46 ` rearnsha at gcc dot gnu dot org
2004-02-09 17:51 ` rearnsha at gcc dot gnu dot org
2004-02-15 12:46 ` gdr at gcc dot gnu dot org
2004-02-22  1:40 ` rearnsha at gcc dot gnu dot org
2004-02-22 16:45 ` gdr at gcc dot gnu dot org
2004-03-01  1:57 ` mmitchel at gcc dot gnu dot org
2004-03-22 21:49 ` mmitchel at gcc dot gnu dot org
2004-03-30 22:22 ` mmitchel at gcc dot gnu dot org
2004-03-31  9:55 ` rearnsha at arm dot com
2004-03-31 17:00 ` mark at codesourcery dot com
2004-04-01 23:30 ` cvs-commit at gcc dot gnu dot org
2004-04-01 23:33 ` mmitchel at gcc dot gnu dot org
2004-04-01 23:33 ` cvs-commit at gcc dot gnu dot org
2004-04-01 23:41 ` cvs-commit at gcc dot gnu dot org
2004-06-12 22:51 ` mmitchel at gcc dot gnu dot org

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).