22fbc67ed0
Don't resolve symbols to PLT entries on the VAX. It just causes a lot of
...
overhead.
2014-03-21 01:40:41 +00:00
f8bf5bbce0
Save actual start of function, not address of entry mask on the stack.
2014-03-19 21:52:00 +00:00
f71aaf7afb
Fix two bugs. (jump past entry mask, fix insv)
2014-03-19 15:34:30 +00:00
db747c73e4
Once we know the bound routine, rebuilt a new callframe that can be unwound
...
properly.
2014-03-19 02:39:22 +00:00
bf06b1032e
Add .cfi ops so that one can unwind through this.
2014-03-18 23:43:38 +00:00
6cb10275d0
Merge riastradh-drm2 to HEAD.
2014-03-18 18:20:35 +00:00
47875e291b
Ignore LD_BIND_NOW for setuid binaries, but keep it in the environment
...
otherwise.
2014-03-18 16:05:34 +00:00
cdbe3bab4f
Change OUTPUT to powerpc:common to make powerpc64 ld happy
2014-03-07 05:57:46 +00:00
9b490ebf92
Make ldd happy
2014-03-07 01:27:14 +00:00
1200e7d3fe
More PPC64 changes.
...
Nothing to do for lazy bindings.
Record DT_PPC64_GLINK and make _rtld_bind return it.
When resolving a JMP_SLOT, copy the source function descriptor into the PLTGOT
2014-03-06 19:19:40 +00:00
15089cc505
Enable building ld.elf_so for powerpc
...
Produce a link map for ld.elf_so
2014-03-06 09:34:56 +00:00
565a18129d
Make ELF64 aware.
2014-03-06 09:34:07 +00:00
fab57a4af5
powerpc64 changes
2014-03-06 07:47:39 +00:00
746070937f
Startup file for powerpc64. (too many difference with ppc32 to do
...
#ifdef dance)
2014-03-06 07:47:19 +00:00
15f3733bef
Catch up with _lwp_park change.
2014-01-31 22:46:40 +00:00
cb2eb1a336
Don't align random sections. Simplify by using .pushsection/.popsection.
2014-01-12 19:36:08 +00:00
467f6149c3
Don't use GOT relocations. Use PC relative for the GOT and GOTOFF for
...
_DYNAMIC. Make thumb friendly.
2013-12-03 00:19:56 +00:00
5fbbc292aa
CID 1132773: Fix file descriptor leak
2013-11-27 18:01:33 +00:00
03ee7fc3bc
If dlopen of the dynamic linker is requested, e.g. by rump in
...
rumpuser_dl_bootstrap, then return &_rtld_objself as the handle and do
not create a duplicate mapping.
The handle is mostly useless as _rtld_objself doesn't appear on
_rtld_objlist.
This fixes a problem on earm platforms where ld.elf_so (currently) has
an init_array section. Calling this caused binaries to segv.
2013-11-20 07:18:23 +00:00
90c0a95fec
Remove hack now it's fixed in 4.8.2
2013-11-16 17:14:06 +00:00
283152d4ee
Apply hack for gcc 4.8 bug on rtld.c.
2013-11-13 11:06:53 +00:00
5e2fc1bbd3
Only initialise TLS space from the PT_TLS segment, if the size is
...
positive and the offset has been computed. Fixes PR lib/48324.
2013-10-21 19:14:15 +00:00
193b880fad
fix unused variable warnings
2013-10-19 17:19:30 +00:00
3ea7549453
Add debug code.
2013-10-19 10:14:06 +00:00
fa741d8823
Use MACHINE_CPU for m68k. Use similar logic for both tests.
2013-10-09 22:33:41 +00:00
abea8b316b
The R_SPARC_L44 relocation type is documented to be truncating and
...
calculated as "(S + A) & 0xfff" - so while it (usually) sets a "imm13" field
in the opcode, only 12 bits of the calculated value can be used.
2013-10-03 10:45:57 +00:00
c60b2e0b87
Missing , in table initializer
2013-09-23 21:00:35 +00:00
25193be039
For earm, emit a MARCH note.
2013-09-10 16:35:10 +00:00
a6c6eb019c
Some architectures can't create unwind tables without the frame pointer,
...
so don't use -fomit-frame-pointer on those.
2013-08-12 17:15:46 +00:00
ffcdb0b170
Fix DEBUG build.
2013-08-03 13:17:05 +00:00
8b8f566d0a
Support MACHINE_ARCH of coldfire
2013-07-18 21:58:13 +00:00
36677b4016
Use new macros from <machine/asm.h> to be more portable
2013-07-18 21:57:42 +00:00
d28e4d0aac
Simplify ELFSIZE by using _MIPS_SZPTR
2013-07-10 15:08:38 +00:00
1f32502b56
Remove redundant assignment. Hasn't been needed since 1999.
2013-05-09 15:47:34 +00:00
c188b90222
remove stray debugging
2013-05-09 15:39:10 +00:00
3c085045e3
convert to SIMPLEQ like the rest of the queues.
2013-05-09 15:38:14 +00:00
a3f195d87d
Fix refcount of the object we found in our list by name.
2013-05-09 15:37:58 +00:00
9b98bab04b
Fix debug symlink
2013-05-09 11:44:36 +00:00
ea2a3471f1
provide debug symlink.
2013-05-08 15:33:45 +00:00
e269a2149a
need xfree() instead of free(). Using free() ends us up in an infinite loop.
2013-05-08 15:25:01 +00:00
558b540714
we don't need the versioning stuff for ldd
2013-05-07 13:01:07 +00:00
5e59dafbb3
don't overwrite COPTS
2013-05-06 22:49:10 +00:00
a3fedff4d1
Fixed handling of DT_SONAME:
...
- add function to add name aliases for shared libraries loaded
XXX[1]: we don't add a name during load time, only when DT_SONAME
is present.
- search already loaded objects in load_by_name for an already
loaded object that matches our name and return it.
- add missing initialization and cleanup for obj->names
XXX[2]: should we make them SIMPLEQ?
- Add XXX in rtld.c about getting the name of an object.
NB: This makes the jdk work again without resorting to a hack of putting
the build path of libjvm.so into the run path (which is a security
problem).
XXX: Pullup-6?
2013-05-06 19:59:29 +00:00
b49eab8ee4
Trailing whitespace
2013-05-06 08:02:20 +00:00
5aa71ff59a
Disalble a debug only where it causes problems.
2013-05-06 07:58:43 +00:00
144fe47c93
Add some DEBUG code.
2013-05-06 07:54:04 +00:00
d631499dcf
Remove dbg for _ctype_
2013-05-03 18:31:24 +00:00
82d2cffc31
Consistently use LDELFSO_MACHINE_ARCH.
2013-05-03 15:52:31 +00:00
653c73bf56
Remove trailing whitespace
2013-05-03 10:27:05 +00:00
558553d0e3
Provide a default LDELFSO_MACHINE_ARCH. hi matt!
...
atf-run works again - the init/fini stuff didn't get run otherwise.
2013-05-03 09:37:17 +00:00
5d1d1ac212
Print out &_DYNAMIC too.
2013-05-02 21:11:18 +00:00
6b1442b03a
Print out PT_DYNAMIC address for objmain.
...
Print out &_DYNAMIC too.
2013-05-02 21:11:03 +00:00
25c44086c9
Changes for MKCOMPAT for earm.
2013-04-27 06:24:22 +00:00
3498abfd33
Use __printflike and __dead instead of __attributes__
2013-04-25 14:16:37 +00:00
0504785998
Use phdr as start of phdrs (duh)
2013-04-25 13:45:15 +00:00
f0744dd4e8
Use ph, not phdr, consistently.
2013-04-25 13:26:17 +00:00
e4c78a6811
Use relocbase, not mapbase.
2013-04-25 13:15:26 +00:00
b469c1cb35
Don't assume that PT_PHDR is always first.
2013-04-25 13:14:41 +00:00
afe1984da4
Adjust exidx_start by relocbase
2013-04-25 12:28:59 +00:00
d4d42c7f62
Handle PT_ARM_EXIDX when mapping objects.
2013-04-25 09:01:23 +00:00
44eef346cd
Only include unwind.h if _RTLD_SOURCE
2013-04-25 06:37:40 +00:00
5417f425b2
Add support for PT_ARM_EXIDX and __gnu_Unwind_Find_exidx for libgcc's support
...
of the ARM EHABI.
2013-04-24 22:37:20 +00:00
e3feeaa73b
Add a ld.elf_so internals version of __gnu_Unwind_Find_exidx for that the
...
libgcc EHABI can find the exception tables for a supplied PC.
2013-04-24 22:32:29 +00:00
f903624708
split debug info for rtld
2013-04-05 20:16:40 +00:00
f7f3399458
don't free cp before we copy it!
2013-01-24 17:57:29 +00:00
f11367c485
Handle CPUFLAGS being empty
2013-01-23 02:31:54 +00:00
69c1ba57d9
Make sure CFLAGS is not empty.
2013-01-23 02:18:50 +00:00
552187152c
If you are using aapcs, you must be using INITFINI
2013-01-22 22:49:09 +00:00
c8b3cc05d4
Fix the other lwp_park() to pass 0 for the 'unpark' lwp.
2012-12-16 23:00:46 +00:00
1ba527af78
You need to pass 0 (not -1) to lwp_park() if you don't also
...
want to do an unpark.
Clearly this code path was never tested!
Needs pullup to netbsd-6
2012-12-16 11:15:21 +00:00
6e8c6cbb85
Add support for R_386_TLS_TPOFF32. This patch was created by
...
Nick Hudson for PR 47061.
2012-11-07 07:24:46 +00:00
9d2c006837
Improve the diagnostic messages for invalid/non-ELF files. Proposed a
...
while back; I forget the context. Has been sitting in one of my trees
awaiting a test run since, apparently, August 2011.
2012-10-13 21:13:07 +00:00
6f24a346a4
On failure in _rtld_dlcheck, call _rtld_error, not xwarnx.
...
Callers need not call _rtld_error themselves now.
Fixes PR lib/41760. Should add the test case from the PR too.
2012-10-01 03:03:46 +00:00
067600c8cf
Revert last, it breaks the tricks xsrc uses.
2012-08-23 21:21:14 +00:00
8e96262375
Remove SHLIBINSTALLDIR and SHLINKINSTALLLIBDIR. Add rpath entries, if
...
SHLIBDIR differs from /usr/lib.
2012-08-23 15:45:02 +00:00
e3dfed33b6
Add .init_array/.fini_array support (conditionalized on HAVE_INITFINI_ARRAY).
2012-08-15 03:46:06 +00:00
4389fd1152
Add a comment for EABI. Use "pop" instructions to save one instruction.
2012-08-13 02:53:25 +00:00
e43469a595
Don't abuse DBG, use COPTS instead.
2012-08-04 15:17:16 +00:00
972dd84db2
Try to make dbg() printfs realy ELFSIZE independent
2012-07-25 22:51:04 +00:00
0180ae1c0a
Add special handling needed for OLO10 relocations.
...
First part of fixing PR port-sparc64/46724.
2012-07-22 09:21:03 +00:00
e97494f3ca
Remove a debug assert that does not hold for PIE (e.g. phdr = 0x40, but
...
obj has not been mapped at 0, so obj->phdr is 0x100040).
OK: skrll
2012-07-22 09:16:35 +00:00
ba562931eb
This version of __diagssert13 is dead.
2012-03-16 11:44:54 +00:00
749108f9b9
Add __diagassert13() so that if libc is compiled with _DIAGNOSTIC, it does
...
not end up bringing in all of stdio.
XXX: This is temporary.
2012-03-15 00:16:07 +00:00
8ae54e9304
Mark _rtld_debug_state as not to be inlined. Add an explicit instruction
...
barrier to prevent removal of calls to it.
2012-03-13 21:00:31 +00:00
52a4c27e3e
Add _dlauxinfo helper function to provide access to the ELF auxilary
...
vector. This can normally be found via __ps_strings, but libc is
initialised too early when linked dynamically and doesn't have access to
it yet, so provide an alternative mechanism via ld.elf_so.
Bump libc minor.
2012-02-16 23:00:38 +00:00
1b54bd1886
Implement lazy binding on hppa. rump_server needs it!?!?!
...
Mostly from OpenBSD.
2012-01-06 10:38:56 +00:00
e86a1635be
Align in the stack before calling _rtld in the way the x86_64 ABI
...
specifies. This is necessary to ensure that init functions are always
called with the correct alignment, e.g. in case they use SSE
instructions.
2011-12-11 22:07:26 +00:00
934a3ac455
Mark xassert as __dead
2011-12-11 11:05:11 +00:00
df83e35987
Remove unused file.
2011-12-05 07:36:32 +00:00
248fe9e5bd
Use the tlsoffset from the correct object.
2011-12-04 16:53:08 +00:00
8737e6c85f
Restore backwards compatibility by removing unnecessary addition of
...
dlvsym to Obj_Entry.
Add some comments.
2011-12-02 09:06:49 +00:00
5b3faf1c05
Unbreak RTLD_NEXT and co on platforms that don't inline or tail call
...
do_dlsym.
2011-11-25 21:27:15 +00:00
7878f85385
Replace if() NADA else if() NADA else continue logic with one explicit
...
block. Split out the MIPS handling to make it explicit and readable.
2011-11-25 14:39:02 +00:00
6133e96c53
Add TLS support for m68k.
2011-11-22 15:25:28 +00:00
10dc456d4b
Hook up TLS support on ARM. Tested by martin@, kernel support from matt@
2011-11-18 16:10:02 +00:00
cc2f98ece2
FreeBSD bug report 161344: TLS area for the main thread is set up to
...
early, if e.g. pointers to functions are used as initializers.
2011-11-17 16:20:11 +00:00
d7bbed6455
avoid linker warnings with the new binutils by not forcing ld.elf_so
...
to be soft-float. the floating-point usage in ld.elf_so is only
because of stdarg functions like printf(), and gcc 4.5 is good
about not actually executing the FP instructions in such functions
unless FP values are passed to them, which ld.elf_so doesn't do.
2011-10-26 15:56:58 +00:00
1f1b5ad097
Don't block SIGTRAP so that on architectures where the debugger needs to
...
be able to receive sigtrap for breakpoints to work, it can. For example
we are setting breakpoints inside dlopen(), after we've blocked the signal,
so the process keep trapping and looping over the trap instruction without
being able to send the signal. Another way would be to move the
_rtld_debug_state() calls outside the critical section...
2011-10-23 21:06:07 +00:00
4a9d1308c3
obsolete libgcc_pic and friends. all libgcc's are supposed to be pic
...
for platforms with shlibs already (and are.)
this obsoletes our hacks for the libgcc specs to use libgcc_pic, and
fixes a couple of other issues reported to me directly.
2011-10-07 09:15:21 +00:00
1862dafe3f
match EXIT() with LEAF_ENTRY() so that sizes are actually constants.
...
fixes binutils 2.21 build issues reported by he and chuq.
2011-09-30 03:05:43 +00:00
matt