Commit Graph

300 Commits

Author SHA1 Message Date
Robert Watson
19ab4b35fb For un-prototyped static inline functions declared in pthread_md.h on
sparc64, use ANSI function headers and specifically indicate the lack of
arguments with 'void'.  Otherwise, warnings are generated at WARNS=3 for
libkse, leading to a compile failure with -Werror.
2007-12-01 14:24:44 +00:00
John Birrell
1566f9a7f9 Include an extra header to get a function prototype. 2007-11-19 08:58:11 +00:00
Marius Strobl
77ddefb873 - Fix the handling of R_SPARC_OLO10, which is a bit of a special case
in the way we implement handling of relocations.
  As for the kernel part this fixes the loading of lots of modules,
  which failed to load due to unresolvable symbols when built after
  the GCC 4.2.0 import. This wasn't due to a change in GCC itself
  though but one of several changes in configuration done along the
  import. Specfically, HAVE_AS_REGISTER_PSEUDO_OP, which causes GCC
  to denote global registers used for scratch purposes and in turn
  GAS uses R_SPARC_OLO10 relocations for, is now defined.
  While at it replace some more ELF_R_TYPE which should have been
  ELF64_R_TYPE_ID but didn't cause problems so far.
- Sync a sanity check between kernel and rtld(1) and change it to be
  maintenance free regarding the type used for the lookup table.
- Sprinkle const on lookup tables.
- Use __FBSDID.

Reported and tested by:	yongari
MFC after:		5 days
2007-10-16 19:17:48 +00:00
Marcel Moolenaar
99227f1ec4 Unbreak the dynamic linker by not creating a cache for rtld-elf
itself. It needs mmap(2), which now needs getosreldate(3) and
which in turn uses a global variable to cache the result. This
cannot be done before linking is done.

See also: ../sparc64/reloc.c:1.15
Approved by: re (kensmith)
2007-07-15 22:52:15 +00:00
Ken Smith
6e5e629f07 Cache does not serve any purpose when rtld is relocating itself, do
not bother allocating one.

Submitted by:	kan
Approved by:	re (bmah)
2007-07-13 16:18:43 +00:00
Alexander Kabaev
03c4fb02dd Add r_debug_state to the list of symbols exported from rtld. GDB needs to
be able to find it in order to trap shared library events from rtld.

Approved by:	re (rwatson)
2007-07-11 23:07:37 +00:00
Christian S.J. Peron
b075897e99 Update the man page to reflect that certain variables will be unset in
the case that the program is set-user-ID or set-group-ID. Add missing
annotations for LIBMAP and LIBMAP_DISABLE.
2007-05-17 19:14:25 +00:00
Christian S.J. Peron
d61e5aa4ed In the event a process is tainted (setuid/setgid binaries), un-set any
potentially dangerous environment variables all together. It should be
noted that the run-time linker will not honnor these environment variables
if the process is tainted currently. However, once a child of the tainted
process calls setuid(2), it's status as being tainted (as defined by
issetugid(2)) will be removed. This could be problematic because
subsequent activations of the run-time linker could honnor these
dangerous variables.

This is more of an anti foot-shot mechanism, there is nothing I am
aware of in base that does this, however there may be third party
utilities which do, and there is no real negative impact of clearing
these environment variables.

Discussed on:	secteam
Reviewed by:	cperciva
PR:		kern/109836
MFC after:	2 weeks
2007-05-17 18:00:27 +00:00
Marcel Moolenaar
cc09c7fb77 Don't enable symbol versioning on ia64 for now. It causes
symbol lookup failures that later result in null-pointer
dereferences. This needs looking into, but since we're
close to release it's possible that it's not resolved before
that time.
2007-05-16 23:24:15 +00:00
Marcel Moolenaar
878dd988ef We don't need --export-dynamic for ld-elf.so.1, because it's a
shared object.
2007-05-16 23:19:51 +00:00
Daniel Eischen
00fb440c1a Enable symbol versioning by default. Use WITHOUT_SYMVER to disable it.
Warning, after symbol versioning is enabled, going back is not easy
(use WITHOUT_SYMVER at your own risk).

Change the default thread library to libthr.

There most likely still needs to be a version bump for at least the
thread libraries.  If necessary, this will happen later.
2007-05-13 14:12:40 +00:00
Pav Lucistnik
88922cf00e Remove %m formatter, it's ifdef 0'ed in the code from the very beginning
MFC after:	1 week
2007-05-12 12:26:25 +00:00
Pav Lucistnik
b9e1d83e61 Expand documentation for LD_TRACE_LOADED_OBJECTS_FMT? variables
PR:		docs/66265 (inspired by)
Submitted by:	Michel Lavondes <fox@vader.aacc.cc.md.us>
MFC after:	1 week
2007-05-12 12:24:49 +00:00
David Xu
25785f9165 Fix a TLS memory leak.
PR: threads/112297
MFC: 1 week
2007-05-05 08:44:59 +00:00
Daniel Eischen
7046601eca Catch up with the private namespace change (s/FBSDprivate/FBSDprivate_1.0). 2007-05-01 13:46:27 +00:00
Alexander Kabaev
effa5b4e71 Retire rtld-specific Versions.def. Symbols exported by rtld are supposed
to override weak symbols exported by libc, so by definition these two
are using the same symbol version names.

Reflect the reality by referring to libc's Versions.def directly.
2007-04-29 16:12:06 +00:00
Daniel Eischen
b5a71dda23 Give the private version namespace a number to match libc. 2007-04-29 14:16:19 +00:00
Daniel Eischen
5f864214bb Use C comments since we now preprocess these files with CPP. 2007-04-29 14:05:22 +00:00
Alexander Kabaev
3c0d0ca74b Bring rtld exports in line with corresponding symbols exported from
libc.

Disable SYMVER_DEFAULT n rtld until its implications are understood
better.
2007-04-09 23:00:29 +00:00
Alexander Kabaev
95a6a22d0a Remove reference to FBSDprivate version. We do not use it in this
module yet.
2007-04-07 23:23:10 +00:00
Alexander Kabaev
ac34654dd9 Catch up on rtld's special status. Since it does not appear on
main object list, its versioning information needs to be examined
separately.

This hopefully fixes problems that people running with SYMVER_ENABLED
are experiencing.
2007-04-07 23:17:00 +00:00
Alexander Kabaev
7ca8e6a670 Prepare rtld for symbol versioning. Disable it by default for now. 2007-04-03 19:01:06 +00:00
Alexander Kabaev
49f90ad282 Implement dl_iterate_phdr function.
Convert boolean flags in internal Obj_Entry structure into bitfields.
Properly check for loaded segment alignment in map_object.
2007-04-03 18:31:20 +00:00
Alexander Kabaev
fb6395da91 Use u_int for variable manipulated by atomic ops to match atomic
ops function prototypes.
2007-04-03 18:28:13 +00:00
John Baldwin
f0ac0f92fb Document LD_UTRACE.
MFC after:	3 days
2007-01-23 22:38:39 +00:00
John Baldwin
1f4b63f824 Add various utrace's for use with ktrace to the ELF runtime linker. To
activate the traces, set the LD_UTRACE (or LD_32_UTRACE) environment
variable.  This also includes code in kdump(8) to parse the traces.

Reviewed by:	kan, jdp
MFC after:	2 weeks
2007-01-09 17:50:05 +00:00
Kip Macy
b84c7a797c Fix TLS on sparc64 for statically and dynamically linked binaries
Approved by: rwatson (mentor)
Reviewed by: jmg and marcel
2006-10-08 02:50:34 +00:00
Jung-uk Kim
da7bf2bb26 Clean up white spaces and fix style(9). 2006-09-19 16:48:08 +00:00
Konstantin Belousov
5fd885b07b Fix the buggy rev. 1.117. dagmembers are only initialized for dlopen'ed
dso that are actually loading. If dso a.so depends on b.so, then dlsym
with handle from dlopen("b.so") will fail unconditionally.

Correct implementation shall use the Obj_Entry.needed list to walk
dependencies DAG.

Test provided by: jkim
Tested (prev. version) by:	jkim, Nicolas Blais <nb_root at videotron ca>, h.blanke at chello nl
Pointy hat to:	kib
Approved by:	kan (mentor)
2006-09-19 12:47:13 +00:00
Ruslan Ermilov
a3ec17db37 Markup fixes. 2006-09-17 21:48:47 +00:00
Konstantin Belousov
d0cb0064fe When looking up the symbol by dlsym, look it not only in the object
given as dso handle, but also in the implicit dependencies of that dso.

Also, const-ify the read-only parameter objlist of symlook_list.

Reported by:	"Simon 'corecode' Schubert" <corecode at fs ei tum de>
Approved by:	kan (mentor)
X-MFC-After:	6.2
2006-09-08 14:59:54 +00:00
Marcel Moolenaar
3614156c7d Fix the variant I allocation for KSE: Allow a larger TCB and assume
that the documented TCB is at the tail of the extended TCB. In other
words, the base of the TCB has a negative offset from the TLS.
2006-09-01 06:13:16 +00:00
Marcel Moolenaar
9d34b1a911 o Fix the static TLS relocation. We were subtracting the size of
the TCB.
o  Use NULL for null pointer argument.
o  Replace magic 8 with TLS_TCB_SIZE.
2006-09-01 06:08:50 +00:00
Marcel Moolenaar
653d825a9a Use NULL for null-pointer argument. 2006-09-01 06:07:26 +00:00
Marcel Moolenaar
862894168d Replace magic 16 with TLS_TCB_SIZE. 2006-08-31 19:42:39 +00:00
Marcel Moolenaar
7d3f85133a Prevent dead code elimination for the TP assignmient by using inline
assembly.
2006-08-30 00:39:07 +00:00
Ruslan Ermilov
2b46c64c9c Remove alpha left-overs. 2006-08-22 08:03:01 +00:00
Xin LI
c93b8edf09 In symlook_obj(): fix _rtld_error output.
MFC After:	2 weeks
2006-08-04 13:37:54 +00:00
Dag-Erling Smørgrav
4421d895a9 *thwack*! all the world's not i386.
Pointy hat to:	des
2006-03-29 12:29:01 +00:00
Dag-Erling Smørgrav
370e5ea48c Don't use dbg if it isn't defined (such as when this file is used by
code outside of rtld-elf)
2006-03-28 18:28:07 +00:00
Dag-Erling Smørgrav
1dc39023f8 Use C99's varadic macro syntax instead of gcc's. 2006-03-28 18:26:47 +00:00
David Xu
c771787169 Er, forgot to clear tls space to zero for Variant II. 2006-03-28 06:14:22 +00:00
David Xu
c0d2338cdd Allocate space for thread pointer, this allows thread library to access
its pointer from begin, and simplifies _get_curthread() in libthr.
2006-03-28 06:09:24 +00:00
Pav Lucistnik
8a50f51b1b - Document LD_32_* environment variables
- Mention 32-bit files in FILES section

MFC after:	1 week
2006-03-27 00:53:58 +00:00
Peter Wemm
8c6a035ead Fix a malloc overrun in 32-bit compat libmap lookup code. 2006-01-31 06:08:28 +00:00
Marcel Moolenaar
05157fa0a1 s/DT_IA64_PLT_RESERVE/DT_IA_64_PLT_RESERVE/ 2006-01-28 17:58:22 +00:00
Marcel Moolenaar
4af16b88cc s/R_IA64_/R_IA_64_/ 2006-01-28 17:56:16 +00:00
Marcel Moolenaar
853b7411b6 s/R_IA64_/R_IA_64_/g as per the ia64 psABI. 2006-01-17 21:03:22 +00:00
Jason Evans
52828c0e9c In preparation for a new malloc implementation:
* Add posix_memalign().

  * Move calloc() from calloc.c to malloc.c.  Add a calloc() implementation in
    rtld-elf in order to make the loader happy (even though calloc() isn't
    used in rtld-elf).

  * Add _malloc_prefork() and _malloc_postfork(), and use them instead of
    directly manipulating __malloc_lock.

Approved by:	phk, markm (mentor)
2006-01-12 07:28:21 +00:00
Alexander Kabaev
96ff9a2bf4 Make lookups for relocations from old unversioned binaries return
oldest versioned symbol available. Do not accept hidden symbols for
all other versions.

Use "<obj->path>: <error message>" for all error messages in new
functions to make them more consistent.
2005-12-24 15:37:30 +00:00