[Bug gold/12879] New: gold doesn't like .symver name,name@VERS
http://sourceware.org/bugzilla/show_bug.cgi?id=12879
Summary: gold doesn't like .symver name,name@VERS
Product: binutils
Version: 2.22 (HEAD)
Status: NEW
Severity: normal
Priority: P2
Component: gold
AssignedTo: i...@airs.com
ReportedBy: rol...@gnu.org
gold gives a multiple definition error when the same object file defines both
foo and foo@... symbols, even with same value. BFD ld accepts this usage.
It would perhaps be reasonable to complain if foo and foo@... did not match.
magilla 100 % head ver.s ver.v
==> ver.s <==
.globl foo
foo:
.symver foo,foo@VERS1
==> ver.v <==
VERS1 {
global:
foo;
};
magilla 101 % gcc -c ver.s
magilla 102 % ./gold/ld-new -shared -o ver.so ver.o --version-script=ver.v
./gold/ld-new: error: ver.o: multiple definition of 'foo'
./gold/ld-new: ver.o: previous definition here
[Exit 1]
magilla 103 % ./ld/ld-new -shared -o ver.so ver.o --version-script=ver.v
magilla 104 % eu-readelf -s ver.o ver.so
ver.o:
Symbol table [ 5] '.symtab' contains 6 entries:
4 local symbols String table: [ 6] '.strtab'
Num:Value Size TypeBind Vis Ndx Name
0: 0 NOTYPE LOCAL DEFAULTUNDEF
1: 0 SECTION LOCAL DEFAULT1
2: 0 SECTION LOCAL DEFAULT2
3: 0 SECTION LOCAL DEFAULT3
4: 0 NOTYPE GLOBAL DEFAULT1 foo
5: 0 NOTYPE GLOBAL DEFAULT1 foo@VERS1
ver.so:
Symbol table [ 2] '.dynsym' contains 6 entries:
1 local symbol String table: [ 3] '.dynstr'
Num:Value Size TypeBind Vis Ndx Name
0: 0 NOTYPE LOCAL DEFAULTUNDEF
1: 0 OBJECT GLOBAL DEFAULT ABS VERS1@@VERS1
2: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS __bss_start
3: 0218 0 NOTYPE GLOBAL DEFAULT5 foo@VERS1
4: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS _edata
5: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS _end
Symbol table [ 8] '.symtab' contains 14 entries:
9 local symbols String table: [ 9] '.strtab'
Num:Value Size TypeBind Vis Ndx Name
0: 0 NOTYPE LOCAL DEFAULTUNDEF
1: 00e8 0 SECTION LOCAL DEFAULT1
2: 0118 0 SECTION LOCAL DEFAULT2
3: 01a8 0 SECTION LOCAL DEFAULT3
4: 01d2 0 SECTION LOCAL DEFAULT4
5: 01e0 0 SECTION LOCAL DEFAULT5
6: 00200218 0 SECTION LOCAL DEFAULT6
7: 00200218 0 OBJECT LOCAL DEFAULT ABS _DYNAMIC
8: 0 OBJECT LOCAL DEFAULT ABS
_GLOBAL_OFFSET_TABLE_
9: 0 OBJECT GLOBAL DEFAULT ABS VERS1
10: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS __bss_start
11: 0218 0 NOTYPE GLOBAL DEFAULT5 foo@VERS1
12: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS _edata
13: 002002f8 0 NOTYPE GLOBAL DEFAULT ABS _end
magilla 105 %
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- 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 gold/12880] New: gold doesn't generate PT_INTERP for .interp section
http://sourceware.org/bugzilla/show_bug.cgi?id=12880 Summary: gold doesn't generate PT_INTERP for .interp section Product: binutils Version: 2.22 (HEAD) Status: NEW Severity: normal Priority: P2 Component: gold AssignedTo: i...@airs.com ReportedBy: rol...@gnu.org In BFD ld, the presence of a regular allocated section called ".interp" causes the implicit generation of a PT_INTERP phdr. This happens under -shared and static links too, and for those (where -dynamic-linker is ignored), it's the available way to get a PT_INTERP (without a manual PHDRS in a linker script). Gold doesn't support this. It treats the .interp section normally as it should, but never emits any PT_INTERP. -- Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email --- 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
