[Bug c/19058] New: setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[Bug c/19058] New: setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[bh at techhouse dot brown dot edu]

In the code that follows, gcc-3.4.1 says:
> gcc -W -Wall  -Wno-unused-parameter qux.c -O2
qux.c:13: warning: variable 'x' might be clobbered by `longjmp' or `vfork'

gcc-4.0 with those same options gives no warning.

Note that neither version warns about 'sum' being clobbered.


#include <signal.h>
#include <setjmp.h>
#include <stdio.h>

static jmp_buf jmpbuf;

static void segv_handler(int sig) {
    longjmp(jmpbuf, 1);
}

int main() {
    int y = 1;
    volatile int *x = &y;
    int sum = 0;

    signal(SIGSEGV, segv_handler);

    if(setjmp(jmpbuf) == 0) {
        while(1) {
            sum += *x;
            x++;
        }
    }

    printf("%d\n", sum);
    return sum;
}

-- 
           Summary: setjmp: 4.0 fails to give 'clobbered' warning
                    (regression from 3.4.1)
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: bh at techhouse dot brown dot edu
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug c/19058] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-12-17 16:18 -------
This is because sum is not used outside of the loop with optimization turned on.  We always use 0.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug tree-optimization/19058] [4.0 Regression] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|setjmp: 4.0 fails to give   |[4.0 Regression] setjmp: 4.0
                   |'clobbered' warning         |fails to give 'clobbered'
                   |(regression from 3.4.1)     |warning (regression from
                   |                            |3.4.1)
   Target Milestone|---                         |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug tree-optimization/19058] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|c                           |tree-optimization
           Keywords|                            |diagnostic


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug tree-optimization/19058] [4.0 Regression] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-01-15 08:24 -------
Hmm, The only thing I can think of is moving this warning to the tree level before we do any 
optimization.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug tree-optimization/19058] [4.0 Regression] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
  GCC build triplet|i686-pc-linux-gnu           |
   GCC host triplet|i686-pc-linux-gnu           |
 GCC target triplet|i686-pc-linux-gnu           |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058

[Bug tree-optimization/19058] [4.0 Regression] setjmp: 4.0 fails to give 'clobbered' warning (regression from 3.4.1)

[rth at gcc dot gnu dot org]

------- Additional Comments From rth at gcc dot gnu dot org  2005-02-02 07:12 -------
The warning has *always* been specifically dependant on which registers are
assigned to which decls.  So it was never assured that you'd get the warning
when you actually needed it.

In this case you're doubly screwed because you expect gcc to magically follow
the path from "*x" through the signal handler, through to setjmp.  Well, we're
not going to do that.

INVALID or WONTFIX, your choice.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19058