[Bug target/37610] New: [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[Bug target/37610] New: [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[danglin at gcc dot gnu dot org]

Executing on host: /home/dave/gcc-4.4/objdir/gcc/testsuite/g++/../../g++
-B/home
/dave/gcc-4.4/objdir/gcc/testsuite/g++/../../
/home/dave/gcc-4.4/gcc/gcc/testsui
te/g++.dg/eh/pr29166.C  -nostdinc++
-I/home/dave/gcc-4.4/objdir/hppa-linux/libst
dc++-v3/include/hppa-linux
-I/home/dave/gcc-4.4/objdir/hppa-linux/libstdc++-v3/i
nclude -I/home/dave/gcc-4.4/gcc/libstdc++-v3/libsupc++
-I/home/dave/gcc-4.4/gcc/
libstdc++-v3/include/backward
-I/home/dave/gcc-4.4/gcc/libstdc++-v3/testsuite/ut
il -fmessage-length=0   -ansi -pedantic-errors -Wno-long-long   
-L/home/dave/gc
c-4.4/objdir/hppa-linux/./libstdc++-v3/src/.libs 
-L/home/dave/gcc-4.4/objdir/hp
pa-linux/./libstdc++-v3/src/.libs
-L/home/dave/gcc-4.4/objdir/hppa-linux/./libib
erty  -lm   -o ./pr29166.exe    (timeout = 300)
PASS: g++.dg/eh/pr29166.C (test for excess errors)
Setting LD_LIBRARY_PATH to
.:/home/dave/gcc-4.4/objdir/hppa-linux/./libstdc++-v3
/src/.libs:/home/dave/gcc-4.4/objdir/hppa-linux/./libstdc++-v3/src/.libs:/home/d
ave/gcc-4.4/objdir/gcc:.:/home/dave/gcc-4.4/objdir/hppa-linux/./libstdc++-v3/src
/.libs:/home/dave/gcc-4.4/objdir/hppa-linux/./libstdc++-v3/src/.libs:/home/dave/
gcc-4.4/objdir/gcc:/home/dave/gcc-4.4/objdir/hppa-linux/libstdc++-v3/.libs:/home
/dave/gcc-4.4/objdir/hppa-linux/libmudflap/.libs:/home/dave/gcc-4.4/objdir/hppa-
linux/libssp/.libs:/home/dave/gcc-4.4/objdir/hppa-linux/libgomp/.libs:/home/dave
/gcc-4.4/objdir/./gcc:/home/dave/gcc-4.4/objdir/./prev-gcc
FAIL: g++.dg/eh/pr29166.C execution test

The program segvs:

Breakpoint 2, 0x402afd30 in _Unwind_DeleteException (exc=<value optimized out>)
    at ../../../gcc/libgcc/../gcc/unwind.inc:276
276         (*exc->exception_cleanup) (_URC_FOREIGN_EXCEPTION_CAUGHT, exc);
Current language:  auto; currently c
(gdb) bt
#0  0x402afd30 in _Unwind_DeleteException (exc=<value optimized out>)
    at ../../../gcc/libgcc/../gcc/unwind.inc:276
#1  0x00012558 in test ()
    at /home/dave/gcc-4.4/gcc/gcc/testsuite/g++.dg/eh/pr29166.C:86
#2  0x00000000 in ?? ()
(gdb) p/x $r22
$2 = 0x16098
(gdb) stepi
0x402afd34      276         (*exc->exception_cleanup)
(_URC_FOREIGN_EXCEPTION_CAUGHT, exc);
(gdb) 
$$dyncall () at ../../../gcc/libgcc/../gcc/config/pa/milli64.S:223
223             bb,>=,n %r22,30,LREF(1)         ; branch if not plabel address
Current language:  auto; currently asm
(gdb) 
223             bb,>=,n %r22,30,LREF(1)         ; branch if not plabel address
(gdb) 
229             bv      %r0(%r22)               ; branch to the real target
(gdb) 
235             stw     %r2,-24(%r30)           ; save return address into
frame marker
(gdb) 
0x00016098 in ?? ()
(gdb) 

Program received signal SIGILL, Illegal instruction.
0x00016098 in ?? ()

This regression was probably introduced in the recent switch to using
assembler CFI directives.


-- 
           Summary: [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution
                    test
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: danglin at gcc dot gnu dot org
 GCC build triplet: hppa-unknown-linux-gnu
  GCC host triplet: hppa-unknown-linux-gnu
GCC target triplet: hppa-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[pinskia at gcc dot gnu dot org]

-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
   Target Milestone|---                         |4.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[mmitchel at gcc dot gnu dot org]

-- 

mmitchel at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[danglin at gcc dot gnu dot org]

------- Comment #1 from danglin at gcc dot gnu dot org  2008-11-18 01:22 -------
*** Bug 38129 has been marked as a duplicate of this bug. ***


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[danglin at gcc dot gnu dot org]

------- Comment #2 from danglin at gcc dot gnu dot org  2008-11-18 02:31 -------
The hppa gas target currently defines DWARF2_LINE_MIN_INSN_LENGTH to 4 in
tc-hppa.h.  I can fix this failure by changing DWARF2_LINE_MIN_INSN_LENGTH
to 1.  However, the failure appears to be caused by an inconsistency in
the handling of DW_CFA_advance_loc1, DW_CFA_advance_loc2 and
DW_CFA_advance_loc4
in dw2gencfi.c and unwind-dw2.c.

In dw2gencfi.c, we have:

            addressT delta = S_GET_VALUE (to) - S_GET_VALUE (from);
            addressT scaled = delta / DWARF2_LINE_MIN_INSN_LENGTH;

            if (scaled <= 0x3F)
              out_one (DW_CFA_advance_loc + scaled);
            else if (delta <= 0xFF)
              {
                out_one (DW_CFA_advance_loc1);
                out_one (delta);
              }
            else if (delta <= 0xFFFF)
              {
                out_one (DW_CFA_advance_loc2);
                out_two (delta);
              }
            else
              {
                out_one (DW_CFA_advance_loc4);
                out_four (delta);
              }

In unwind-dw2.c, we have:

        case DW_CFA_advance_loc1:
          fs->pc += read_1u (insn_ptr) * fs->code_align;
          insn_ptr += 1;
          break;
        case DW_CFA_advance_loc2:
          fs->pc += read_2u (insn_ptr) * fs->code_align;
          insn_ptr += 2;
          break;
        case DW_CFA_advance_loc4:
          fs->pc += read_4u (insn_ptr) * fs->code_align;
          insn_ptr += 4;
          break;

The code in binutils/dwarf.c also scales the file values by the code
alignment factor.  So two against one says the dw2gencfi.c code is wrong.


-- 

danglin at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rth at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #3 from jakub at gcc dot gnu dot org  2008-11-18 09:03 -------
libunwind also scales all of DW_CFA_advance_loc{,1,2,4} by code alignment
factor and IMHO it matches the standard wording too:
DW_CFA_advance_loc
  The DW_CFA_advance instruction takes a single operand (encoded with the
opcode) that represents a constant delta. The required action is to create a
new table row with a location value that is computed by taking the current
entry's location value and adding the value of delta * code_alignment_factor.
All other values in the new row are initially identical to the current row.
DW_CFA_advance_loc1
  The DW_CFA_advance_loc1 instruction takes a single ubyte operand that
represents a constant delta. This instruction is identical to
DW_CFA_advance_loc except for the encoding and size of the delta operand.

Can you submit the trivial fix (replace 3 times (delta) with (scaled) in
dw2gencfi.c to binutils?  On the GCC side we'll need to write a configury test
which will fail if this bug is detected and the target has code alignment
factor bigger than 1.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #4 from jakub at gcc dot gnu dot org  2008-11-18 09:19 -------
On the GCC configury side (and in gas testsuite too) it is just a matter of:
cat > conftest.s <<EOF
.cfi_startproc
.cfi_adjust_cfa_offset 64
.skip 512, 0
.cfi_adjust_cfa_offset 128
.cfi_endproc
EOF
$gcc_cv_as -o conftest.o conftest.s
if $gcc_cv_readelf -wf conftest.o 2>/dev/null | egrep -q
'DW_CFA_advance_loc2:[[:blank:]]*512[[:blank:]]*'; then
  gcc_cv_as_cfi_advance_working=yes;
else
  gcc_gc_as_cfi_advance_working=no;
fi


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[dave at hiauly1 dot hia dot nrc dot ca]

------- Comment #5 from dave at hiauly1 dot hia dot nrc dot ca  2008-11-18 14:06 -------
Subject: Re:  [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

> Can you submit the trivial fix (replace 3 times (delta) with (scaled) in
> dw2gencfi.c to binutils?  On the GCC side we'll need to write a configury test
> which will fail if this bug is detected and the target has code alignment
> factor bigger than 1.

Will submit the dw2gencfi.c fix.   Tested last night.

Dave


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[danglin at gcc dot gnu dot org]

------- Comment #6 from danglin at gcc dot gnu dot org  2008-11-18 14:56 -------
The binutils patch is here:
http://sourceware.org/ml/binutils/2008-11/msg00200.html


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[ebotcazou at gcc dot gnu dot org]

------- Comment #7 from ebotcazou at gcc dot gnu dot org  2008-11-18 18:48 -------
*** Bug 38131 has been marked as a duplicate of this bug. ***


-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |laurent at guerby dot net


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[ebotcazou at gcc dot gnu dot org]

------- Comment #8 from ebotcazou at gcc dot gnu dot org  2008-11-18 18:51 -------
Likewise on SPARC.


-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ebotcazou at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2008-11-18 18:51:34
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #9 from jakub at gcc dot gnu dot org  2008-11-29 21:05 -------
The binutils fix has been checked in, I'll work on configury on Monday.


-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |jakub at gcc dot gnu dot org
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2008-11-18 18:51:34         |2008-11-29 21:05:42
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #10 from jakub at gcc dot gnu dot org  2008-12-02 15:48 -------
Created an attachment (id=16811)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16811&action=view)
gcc44-pr37610.patch

Patch I've briefly tested on ppc-linux with broken as well as fixed gas and on
x86_64-linux am bootstrapping it ATM.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #11 from jakub at gcc dot gnu dot org  2008-12-03 09:11 -------
Subject: Bug 37610

Author: jakub
Date: Wed Dec  3 09:09:43 2008
New Revision: 142389

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=142389
Log:
        PR target/37610
        * configure.ac (gcc_cv_readelf): Look for readelf.
        (gcc_cv_as_cfi_advance_working): Check for working
        cfi advances with code alignment factor > 1.
        (HAVE_GAS_CFI_DIRECTIVE): Don't define if cfi advances
        don't work properly.
        * configure: Regenerated.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/configure
    trunk/gcc/configure.ac


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[jakub at gcc dot gnu dot org]

------- Comment #12 from jakub at gcc dot gnu dot org  2008-12-03 09:11 -------
Fixed in gas, configury check added for gcc.


-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610

[Bug target/37610] [4.4 Regression] FAIL: g++.dg/eh/pr29166.C execution test

[ebotcazou at gcc dot gnu dot org]

------- Comment #13 from ebotcazou at gcc dot gnu dot org  2008-12-20 22:33 -------
Subject: Bug 37610

Author: ebotcazou
Date: Sat Dec 20 22:32:30 2008
New Revision: 142850

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=142850
Log:
        PR target/37610
        * configure.ac (gcc_cv_as_cfi_advance_working): Fall back to 'no'
        if readelf is nowhere to be found.
        * configure: Regenerate.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/configure
    trunk/gcc/configure.ac


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37610