[Bug rtl-optimization/65064] New: [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[Bug rtl-optimization/65064] New: [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

            Bug ID: 65064
           Summary: [5.0 regression] FAIL: gcc.dg/torture/pr60115.c   -O1
                    (test for excess errors)
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: schwab@linux-m68k.org
                CC: hjl at gcc dot gnu.org, rth at gcc dot gnu.org
            Target: ia64-*-*

$ gcc/xgcc -B gcc/ ../gcc/testsuite/gcc.dg/torture/pr60115.c -O
/usr/ia64-suse-linux/bin/ld: a.out: __gp does not cover short data segment
collect2: error: ld returned 1 exit status

Broken by r220674.

--- gcc-20150213/Build/pr60115.s        2015-02-14 21:36:37.000000000 +0100
+++ gcc-20150214/Build/pr60115.s        2015-02-14 21:35:30.000000000 +0100
@@ -7,22 +7,20 @@
        .proc main#
 main:
        .prologue
-       .save ar.lc, r3
-       mov r3 = ar.lc
+       .save ar.lc, r2
+       mov r2 = ar.lc
        .body
 [.L2:]
-       addl r14 = @ltoffx(a#), r1
-       ;;
-       ld8.mov r14 = [r14], a#
+       addl r14 = @gprel(a#), gp
        ;;
        ld4 r14 = [r14]
        ;;
        cmp4.eq p6, p7 = 0, r14
-       addl r2 = @ltoffx(b#), r1
+       addl r16 = @ltoffx(b#+32768), r1
        ;;
-       ld8.mov r2 = [r2], b#
+       ld8.mov r16 = [r16], b#+32768
        ;;
-       addl r16 = 40000, r2
+       adds r16 = 7232, r16
 .L4:
        .pred.rel.mutex p6, p7
        (p6) br.cond.dpnt .L3
@@ -40,17 +38,15 @@
        adds r14 = -1, r14
        ;;
        br.cloop.sptk.few .L6
-       addl r14 = @ltoffx(a#), r1
-       ;;
-       ld8.mov r14 = [r14], a#
+       addl r14 = @gprel(a#), gp
        ;;
        st4 [r14] = r0
 .L3:
        mov r8 = r0
-       mov ar.lc = r3
+       mov ar.lc = r2
        br.ret.sptk.many b0
        ;;
        .endp main#
        .common b#,8,4
        .common a#,4,4
-       .ident  "GCC: (GNU) 5.0.0 20150213 (experimental) [trunk revision
220673]"
+       .ident  "GCC: (GNU) 5.0.0 20150214 (experimental) [trunk revision
220705]"

[Bug rtl-optimization/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2015-02-15
     Ever confirmed|0                           |1

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> ---
Can you provide all glibc input files to ld, including crt*.o,
ld-linux-ia64.so.2, libc.so, ..., so that I can reproduce it
with a cross linker on Linux/x86-64?

[Bug rtl-optimization/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> ---
Does -fno-common fix the problem?  If it is the case, we need to
treat common symbol differently for ia64 since

    .common    b#,8,4
    .common    a#,4,4

won't be placed in small data section reachable via GP reloc. The
problem for ia64 is even if a data symbol is defined locally, we
can't use GP reloc if it isn't placed in small data section.

[Bug rtl-optimization/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> ---
Created attachment 34759
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34759&action=edit
A patch

Please try this.

[Bug target/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

Andreas Schwab <schwab@linux-m68k.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
          Component|rtl-optimization            |target

--- Comment #4 from Andreas Schwab <schwab@linux-m68k.org> ---
This fixes the test case at all opt levels.

[Bug target/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #34759|0                           |1
        is obsolete|                            |

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> ---
Created attachment 34766
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34766&action=edit
An updated patch

Please try this one instead.

[Bug target/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #6 from Andreas Schwab <schwab@linux-m68k.org> ---
This works as well.

[Bug target/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #7 from H.J. Lu <hjl.tools at gmail dot com> ---
A patch is posted at

https://gcc.gnu.org/ml/gcc-patches/2015-02/msg00913.html

[Bug target/65064] [5.0 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #8 from Andreas Schwab <schwab@linux-m68k.org> ---
It doesn't bootstrap though:

ada/ali.o: In function `ali__alis__reallocate':
ali.adb:(.text+0x2170): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__units__reallocate':
ali.adb:(.text+0x2f00): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__interrupt_states__reallocate':
ali.adb:(.text+0x4150): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__specific_dispatching__reallocate':
ali.adb:(.text+0x5540): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__withs__reallocate':
ali.adb:(.text+0x6a00): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__args__reallocate':
ali.adb:(.text+0x8080): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__linker_options__reallocate':
ali.adb:(.text+0x92b0): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__notes__reallocate':
ali.adb:(.text+0xa570): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__no_deps__reallocate':
ali.adb:(.text+0xc6b0): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__sdep__reallocate':
ali.adb:(.text+0xd850): relocation truncated to fit: GPREL22 against `.rodata'
ada/ali.o: In function `ali__xref_section__reallocate':
ali.adb:(.text+0xf3a0): additional relocation overflows omitted from the output
/usr/ia64-suse-linux/bin/ld: final link failed: Nonrepresentable section on
output
collect2: error: ld returned 1 exit status

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |5.0
            Summary|[5.0 regression] FAIL:      |[5 regression] FAIL:
                   |gcc.dg/torture/pr60115.c    |gcc.dg/torture/pr60115.c
                   |-O1  (test for excess       |-O1  (test for excess
                   |errors)                     |errors)

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #9 from Andreas Schwab <schwab@linux-m68k.org> ---
before:
    2170:       02 38 01 02 00 24       [MII]       addl r39=0,r1
                        2170: GPREL22   .sdata+0x38

after:
    2170:       02 38 01 02 00 24       [MII]       addl r39=0,r1
                        2170: GPREL22   .rodata+0x114

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

Andreas Schwab <schwab@linux-m68k.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW

--- Comment #11 from Andreas Schwab <schwab@linux-m68k.org> ---
Ada doesn't preprocess.

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #12 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Andreas Schwab from comment #11)
> Ada doesn't preprocess.

Can you try my first patch? If it works, I can explain why :-(.

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #13 from Andreas Schwab <schwab@linux-m68k.org> ---
This has the same problem.

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #14 from Andreas Schwab <schwab@linux-m68k.org> ---
I've accidentally tested with both patches applied, will try again with only
the first one.

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[schwab@linux-m68k.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #15 from Andreas Schwab <schwab@linux-m68k.org> ---
The first patch works without regressions.

http://gcc.gnu.org/ml/gcc-testresults/2015-02/msg02066.html

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #16 from H.J. Lu <hjl.tools at gmail dot com> ---
A new patch is posted at

https://gcc.gnu.org/ml/gcc-patches/2015-02/msg01105.html

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

--- Comment #17 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> ---
Author: hjl
Date: Wed Feb 18 17:24:20 2015
New Revision: 220792

URL: https://gcc.gnu.org/viewcvs?rev=220792&root=gcc&view=rev
Log:
Return false for common symbols in sdata_symbolic_operand

Although common symbols are defined in executables, they aren't in small
data section.  But a definition in small data section overrides a common
symbol, which still binds lcoally, and turns a reference to common symbol
to reference to small data section.  Even if ia64_in_small_data_p returns
true on common symbols, sdata_symbolic_operand must return false on common
symbols.  Common symbols are assumed to be placed in small data section,
but are accessed as if they are in normal data section so that they won't
cause any relocation overflow.

    PR target/65064
    * config/ia64/predicates.md (sdata_symbolic_operand): Return false
    for common symbols.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/ia64/predicates.md

[Bug target/65064] [5 regression] FAIL: gcc.dg/torture/pr60115.c -O1 (test for excess errors)

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65064

H.J. Lu <hjl.tools at gmail dot com> changed:

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

--- Comment #18 from H.J. Lu <hjl.tools at gmail dot com> ---
Fixed.