[Bug ld/16563] Corrupt .eh-frame section created when linking LTO and non-LTO objects

[nickc at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16563

Nick Clifton  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--- Comment #9 from Nick Clifton  ---
Hi Alan,

  Thanks for the patch - it does indeed solve the problem.

  I believe however that readelf is/was correct in saying the CIE pointer was
invalid.  These pointers are offsets into the .debug_frame section of the CIE
entry to be used by the FDE that is being decoded.  But the offsets can never
be negative (ie pointing to before the start of .debug_frame), so readelf was
correct to complain with this particular testcase.

  You are correct however in saying that readelf does have a potential bug - it
will complain about FDEs that reference a CIE that is defined later on in the
.debug_frame section.  But I have never yet seen a binary that does this.  When
one turns up I will fix readelf. :-)

Cheers
  Nick

-- 
You are receiving this mail because:
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils

[Bug ld/17277] New: ld -shared: bogus DT_TEXTREL marker (and R_ARM_NONE) for PC-relative cross-section relocs (e.g. movw/movt)

[roland at gnu dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=17277

Bug ID: 17277
   Summary: ld -shared: bogus DT_TEXTREL marker (and R_ARM_NONE)
for PC-relative cross-section relocs (e.g. movw/movt)
   Product: binutils
   Version: 2.24
Status: NEW
  Severity: normal
  Priority: P2
 Component: ld
  Assignee: unassigned at sourceware dot org
  Reporter: roland at gnu dot org

Reproduced with:

$ cat pcrel-shared.s
.syntax unified
.arm
.arch armv7-a

.text
.globl foo
.type foo,%function
foo:movw r0, #:lower16:symbol - 1f - 8
movt r0, #:upper16:symbol - 1f - 8
1:add r0, pc
bx lr

.data
.globl symbol
.hidden symbol
symbol:.long 22
$ arm-linux-gnueabi-as -o pcrel-shared.o pcrel-shared.s
$ arm-linux-gnueabi-ld -shared -o pcrel-shared.so pcrel-shared.o
$ readelf -dr pcrel-shared.so

Dynamic section at offset 0x1e4 contains 10 entries:
  TagType Name/Value
 0x0004 (HASH)   0x94
 0x0005 (STRTAB) 0x184
 0x0006 (SYMTAB) 0xd4
 0x000a (STRSZ)  63 (bytes)
 0x000b (SYMENT) 16 (bytes)
 0x0011 (REL)0x1c4
 0x0012 (RELSZ)  16 (bytes)
 0x0013 (RELENT) 8 (bytes)
 0x0016 (TEXTREL)0x0
 0x (NULL)   0x0

Relocation section '.rel.dyn' at offset 0x1c4 contains 2 entries:
 Offset InfoTypeSym.Value  Sym. Name
   R_ARM_NONE   
   R_ARM_NONE   
$

The R_ARM_NONE relocs are bogus.  More important is the DT_TEXTREL marker,
which is harmfully bogus.

Appears in 2.24 (and earlier) and trunk du jour.  Gold does not have this bug.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils

[Bug binutils/17278] New: ld does not accept UTF8-encoded response files ('@file') on windows using mingw

[rafaelauler at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17278

Bug ID: 17278
   Summary: ld does not accept UTF8-encoded response files
('@file') on windows using mingw
   Product: binutils
   Version: 2.24
Status: NEW
  Severity: normal
  Priority: P2
 Component: binutils
  Assignee: unassigned at sourceware dot org
  Reporter: rafaelauler at gmail dot com

When using response files with international characters on mingw/windows, I
must encode then in the old/legacy ISO-8859-1 or whatever is the system's
current code page. There is no support for UTF8 with/without BOM or UTF16.

This is weird since on UNIX/cygwin hosts, the response files use UTF8. This
means that ld for mingw and ld for cygwin cannot share the same response files.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils