Re: target/6040: [ARM] interrupt attribute generats wrong code (3.1-cvs)

Re: target/6040: [ARM] interrupt attribute generats wrong code (3.1-cvs)

[Dara Hazeghi]

The following reply was made to PR target/6040; it has been noted by GNATS.

From: Dara Hazeghi <dhazeghi@yahoo.com>
To: tori@unhappy.mine.nu, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Cc:  
Subject: Re: target/6040: [ARM] interrupt attribute generats wrong code (3.1-cvs)
Date: Fri, 16 May 2003 19:59:20 -0700

 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit- 
 trail&database=gcc&pr=6040
 
 Hello,
 
 here's another bug against a fairly old version of gcc. Gcc 3.3  
 produces:
          .file   "junk.c"
          .text
          .align  2
          .global foo
          .type   foo, %function
 foo:
          @ Interrupt Service Routine.
          @ args = 0, pretend = 0, frame = 4
          @ frame_needed = 1, uses_anonymous_args = 0
          str     ip, [sp, #-4]!
          mov     ip, sp
          sub     lr, lr, #4
          stmfd   sp!, {r3, fp, ip, lr, pc}
          sub     fp, ip, #4
          sub     sp, sp, #4
          mov     r3, #0
          str     r3, [fp, #-24]
          ldmea   fp, {r3, fp, ip, pc}^
          .size   foo, .-foo
          .ident  "GCC: (GNU) 3.3 20030508 (prerelease)"
 
 with gcc mainline:
          .file   "junk.c"
          .text
          .align  2
          .global foo
          .type   foo, %function
 foo:
          @ Interrupt Service Routine.
          @ args = 0, pretend = 0, frame = 4
          @ frame_needed = 1, uses_anonymous_args = 0
          str     ip, [sp, #-4]!
          mov     ip, sp
          stmfd   sp!, {r3, fp, ip, lr, pc}
          sub     fp, ip, #4
          sub     sp, sp, #4
          mov     r3, #0
          str     r3, [fp, #-24]
          ldmea   fp, {r3, fp, sp, lr}
          ldmfd   sp!, {ip}
          subs    pc, lr, #4
          .size   foo, .-foo
          .ident  "GCC: (GNU) 3.4 20030508 (experimental)"
 
 can you determine if this is correct? Thanks,
 
 Dara
 

Re: target/6040: ARM interrupt attribute generats wrong code (3.1-cvs)

[mmitchel]

Synopsis: ARM interrupt attribute generats wrong code (3.1-cvs)

State-Changed-From-To: open->analyzed
State-Changed-By: mmitchel
State-Changed-When: Thu Apr 18 09:04:02 2002
State-Changed-Why:
    Confirmed.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=6040

target/6040: ARM interrupt attribute generats wrong code (3.1-cvs)

[tori]

>Number:         6040
>Category:       target
>Synopsis:       ARM interrupt attribute generats wrong code (3.1-cvs)
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Mar 22 08:06:01 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Tobias Ringstrom
>Release:        gcc version 3.1 20020321 (prerelease)
>Organization:
>Environment:
Crosscompiler for arm-elf
>Description:
Compile the following code with "arm-elf-gcc -S"

void foo(void) __attribute__((__interrupt__));
void foo()
{
        int a=0;
}

gives the following assembler code:

        .file   "tmp2.c"
        .text
        .align  2
        .global foo
        .type   foo,function
foo:
        @ Interrupt Service Routine.
        @ args = 0, pretend = 0, frame = 4
        @ frame_needed = 1, uses_anonymous_args = 0
        str     ip, [sp, #-4]!
        mov     ip, sp
        sub     lr, lr, #4
        stmfd   sp!, {r3, fp, ip, lr, pc}
        sub     fp, ip, #4
        sub     sp, sp, #4
        mov     r3, #0
        str     r3, [fp, #-24]
        ldmea   fp, {r3, fp, ip, pc}^
.Lfe1:
        .size   foo,.Lfe1-foo
        .ident  "GCC: (GNU) 3.1 20020321 (prerelease)"

Major problem:
The stack pointer (sp) is never restored!

Comment/question:
There are four lost bytes, but that might be due to alignment of the stack.  Is it?

/Tobias
>How-To-Repeat:
See Description above!
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: