relocations until tls is initialized and stacks permissions correctly
set. This allows the ifunc to call malloc(3) and some other heavy
services.
Add debug banner.
MFC after: 3 days
from the dispatcher would also acquire bind lock in read mode, which
is the supported operation. plt is explicitely designed to allow safe
multithreaded updates, so the shared lock do not cause problems.
The error in r228435 is that it allows read lock acquisition after the
write lock for the bind block. If we dlopened the shared object that
contains IRELATIVE or jump slot which target is STT_GNU_IFUNC, then
possible recursive plt resolve from the dispatcher would cause it.
Postpone the resolution for irelative/ifunc right before initializers
are called, and drop bind lock around calls to dispatcher. Use
initlist to iterate over the objects instead of the ->next, due to
drop of the bind lock in iteration.
For i386/reloc.c:reloc_iresolve(), fix calculation of the dispatch
function address for dso, by taking into account possible non-zero
relocbase.
MFC after: 3 weeks
rtld on 386 and amd64. This adds runtime bits neccessary for the use
of the dispatch functions from the dynamically-linked executables and
shared libraries.
To allow use of external references from the dispatch function, resolution
of the R_MACHINE_IRESOLVE relocations in PLT is postponed until GOT entries
for PLT are prepared, and normal resolution of the GOT entries is finished.
Similar to how it is done by GNU, IRELATIVE relocations are resolved in
advance, instead of normal lazy handling for PLT.
Move the init_pltgot() call before the relocations for the object are
processed.
MFC after: 3 weeks
relocations are processed, since tls initialization section might be
itself subject for relocations. Only set up of the block is postponed,
the tls block offsets are allocated before relocation processing, since
TLS-related relocations may need offsets ready.
Reported by: ale
PR: threads/161344
Reviewed by: kan
MFC after: 1 week
function (a hook necessary for gdb support), is inlined, but since the
function contains no code, no calls to it are generated. When gdb is
debugging a dynamically linked program, this causes backtraces to be
corrupted.
Fix it by marking the function __noinline, and inserting an empty asm
statement, that pretends to clobber memory. This forces the compiler to
emit calls to r_debug_state() throughout rtld.c.
Approved by: re (kib)
C runtime services, like printf(). Unfortunately, the multithread-safeness
measures in the libc do not work in rtld environment.
Rip the kernel printf() implementation and use it in the rtld instead of
libc version. This printf does not require any shared global data and thus
is mt-safe. Systematically use rtld_printf() and related functions, remove
the calls to err(3).
Note that stdio is still pulled from libc due to libmap implementaion using
fopen(). This is safe but unoptimal, and can be changed later.
Reported and tested by: pgj
Diagnosed and reviewed by: kan (previous version)
Approved by: re (bz)
The second close(2) call resulted in heisenbugs in some multi-threaded
applications where e.g. dlopen(3) call in one thread could close a file
descriptor for a file having been opened in other thread concurrently.
My litmus test for this issue was an openoffice.org build.
Reviewed by: jhb
MFC after: 2 weeks
Instead of aborting in locate_dependency(), propagate the error to
caller. The rtld startup function does the right thing with an error
from rtld_verify_versions(), depending on the mode of operation.
Reported by: maho
In collaboration with: kan
MFC after: 1 week
Place elements on DAG lists in breadth-first order. This allows us to
walk pre-built list in all cases where breadth-first dependency chain
enumeration is required.
Fix dlsym on special handle obtained by dlopen(NULL, ...) to do what
comment claims it does. Take advantage of recently added symlook_global
function to iterate over main objects and global DAGs lists properly in
search of a symbol. Since rtld itself provides part of the global
namespace, search rtld_obj too.
Remove recursion from init_dag and symlook_needed functions. Use
symlook_needed for ELF filtee processing only and change lookup order
used in the function to match the order used by Solaris runtime linker
under same circumstances. While there, fix weak symbol handling in the
loop so that we return the first weak symbol definition if no strong one
was found, instead of the last one.
Reviewed by: kib
MFC after: 1 month
one. Search global objects, together with main object and
dependencies, for the requested symbol.
Move the common code from symlook_default() into new helper
symlook_global(), and use it both in symlook_global() and
get_program_var_addr().
Supply lock state to get_program_var_addr().
Reviewed by: kan
Tested by: Mykola Dzham <i levsha me>
by kernel, and parse PT_GNU_STACK phdr from linked and loaded dsos.
If the loaded dso requires executable stack, as specified by PF_X bit
of p_flags of PT_GNU_STACK phdr, but current stack protection does not
permit execution, the __pthread_map_stacks_exec symbol is looked up
and called. It should be implemented in libc or threading library and
change the protection mode of all thread stacks to be executable.
Provide a private interface _rtld_get_stack_prot() to export the stack
access mode as calculated by rtld.
Reviewed by: kan
filters are implemented.
Filtees are loaded on demand, unless LD_LOADFLTR environment variable
is set or -z loadfltr was specified during the linking. This forces
rtld to upgrade read-locked rtld_bind_lock to write lock when it
encounters an object with filter during symbol lookup.
Consolidate common arguments of the symbol lookup functions in the
SymLook structure. Track the state of the rtld locks in the
RtldLockState structure. Pass local RtldLockState through the rtld
symbol lookup calls to allow lock upgrades.
Reviewed by: kan
Tested by: Mykola Dzham <i levsha me>, nwhitehorn (powerpc)
dlclose() call may unload the object of the outer call prematurely
because objects are unreferenced before _fini() calls.
Fix this by unreferencing objects after calling objlist_call_fini() in
dlclose(). Therefore objlist_call_fini() now calls the fini function if
the reference count of an object is 1. In addition we must restart the
list_fini traversal after every _fini() call because another dlclose()
call might have modified the reference counts.
Add an XXX comment to objlist_call_fini() about possible race with
dlopen().
PR: 133246, 149464
Reviewed by: kan, kib
loaded as a dependency and marked -z nodlopen, object' DAG is already
initialized by load_needed_objects(). Due to this, the init_dag() call
from dlopen() does not increment refcount for the object [1].
Change init_dag() to not increment DAG refcount. Require explicit calls
to ref_dag() to increment, and assert that ref_dag() and unref_dag()
are called for root that has dag initialized. To fix the noted issue,
unconditionally call both init_dag() and ref_dag() in dlopen() for the
case when the object was already loaded, making it similar to the case
of newly loaded object.
Noted by: jh [1]
Reviewed by: jh, kan
MFC after: 6 days
dependency, then the dso never has its DAG initialized. Empty DAG
makes ref_dag() call in dlopen() a nop, and the dso refcount is off
by one.
Initialize the DAG on the first dlopen() call, using a boolean flag
to prevent double initialization.
From the PR (edited):
Assume we have a library liba.so, containing a function a(), and a
library libb.so, containing function b(). liba.so needs functionality
from libb.so, so liba.so links in libb.so.
An application doesn't know about the relation between these libraries,
but needs to call a() and b(). It dlopen()s liba.so and obtains a
pointer to a(), then it dlopen()s libb.so and obtains a pointer to b().
As soon as the application doesn't need a() anymore, it dlclose()s liba.so.
Expected result: the pointer to b() is still valid and can be called
Actual result: the pointer to b() has become invalid, even though the
application did not dlclose() the handle to libb.so. On calling b(), the
application crashes with a segmentation fault.
PR: misc/151861
Based on patch by: jh
Reviewed by: kan
Tested by: Arjan van Leeuwen <freebsd-maintainer opera com>
MFC after: 1 week
atexit and __cxa_atexit handlers that are either installed by unloaded
dso, or points to the functions provided by the dso.
Use _rtld_addr_phdr to locate segment information from the address of
private variable belonging to the dso, supplied by crtstuff.c. Provide
utility function __elf_phdr_match_addr to do the match of address against
dso executable segment.
Call back into libthr from __cxa_finalize using weak
__pthread_cxa_finalize symbol to remove any atfork handler which
function points into unloaded object.
The rtld needs private __pthread_cxa_finalize symbol to not require
resolution of the weak undefined symbol at initialization time. This
cannot work, since rtld is relocated before sym_zero is set up.
Idea by: kan
Reviewed by: kan (previous version)
MFC after: 3 weeks
Use local version of getpagesize(), rtld_getpagesize() in private allocator.
Override the __getosreldate() previously fetched from libc_pic.a with
local version that uses aux value if present. Note that __getosreldate()
is used by rtld indirectly, by mmap(2) libc wrapper.
To be able to utilize aux, split digest_dynamic() for use by init_rtld()
into two parts, where the first one does not call malloc(), and the
second part uses it. init_rtld() is able to initialize global variables
before digest_dynamic2() calls. In particular, pagesize and osreldate are
set up from the aux values.
Now, rtld avoids (two) sysctl calls in startup.
Tested by: marius (sparc64)
MFC after: 1 month
fix for it.
The bug occurs when using the --as-needed flag to ld in the presence of
synthetic linker-generated symbols that reference symbols defined in
linked-to shared libraries with versioned symbols. When the only symbols
used from a library fall into this category, ld will drop the DT_NEEDED
entry for it, but retain the versioning information. This bug is best
fixed/hacked around in binutils, not in rtld.
Discussed with: kan
because calling mmap() etc. may use GOT which is not set up
yet. Use calloc() instead of mmap() in cases where this
was the case before (sparc64, powerpc, arm).
Submitted by: Dimitry Andric (dimitry andric com)
Reviewed by: kan
Approved by: ed (mentor)
o Use obj->tlsinitsize to determine whether there's initialized data.
o If obj->tlssize > obj->tlsinitsize, then bzero uninitialized data.
o Don't exclude variant I from the work-around in free_tls_offset().
adjustment for all virtual addresses encoded into the ELF structures of
it. PIE binary could and should be loaded at non-zero mapbase.
For sym_zero pseudosymbol used as a return value from find_symdef()
for undefined weak symbols, st_value also should be adjusted, since
_rtld_bind corrects symbol values by relocbase.
Discussed with: bz
Reviewed by: kan
Tested by: bz (i386, amd64), bsam (linux)
MFC after: some time
altered through their .init code. This might happen if init
vector calls dlopen on its own and that dlopen causes some not
yet initialized object to be initialized earlier as part of that
dlopened DAG.
Do not reset module reference counts to zero on final fini vector
run when process is exiting. Just add an additional parameter to
force fini vector invocation regardless of current reference count
value if object was not destructed yet. This allows dlclose called
from fini vector to proceed normally instead of failing with handle
validation error.
Reviewed by: kib
Reported by: venki kaps
dlsym seaches using this handle are expected to look for symbol
definitions in all objects loaded at the program start time along
with all objects currently in RTLD_GLOBAL scope.
Discussed with: kib
Reported by: Maho NAKATA
MFC after: 2 weeks
dlfunc() called dlsym() to do the work, and dlsym() determines the dso
that originating the call by the return address. Due to this, dlfunc()
operated as if the caller is always the libc.
To fix this, move the dlfunc() to rtld, where it can call the internal
implementation of dlsym, and still correctly fetch return address.
Provide usual weak stub for the symbol from libc for static binaries.
dlfunc is put to FBSD_1.0 symver namespace in the ld.so export to
override dlfunc@FBSD_1.0 weak symbol, exported by libc.
Reported, analyzed and tested by: Tijl Coosemans <tijl ulyssis org>
PR: standards/133339
Reviewed by: kan
the rtld hints file. This environment variable would be unset if the
process is considered as tainted with setuid/setgid. This feature gives
a convenient way of using a custom set of shared library that is not
located in the default location and switch back.
Feature requested by: iXsystems
Original patch by: John Hixson
MFC after: 2 weeks
soneeded pathes. The $ORIGIN, $OSNAME, $OSREL and $PLATFORM tokens
are supported. Enabling the substitution requires DF_ORIGIN flag in
DT_FLAGS or DF_1_ORIGIN if DF_FLAGS_1, that may be set with -z origin
gnu ld flag. Translation is unconditionally disabled for setuid/setgid
processes.
The $ORIGIN translation relies on the AT_EXECPATH auxinfo supplied
by kernel.
Requested by: maho
Tested by: maho, pho
Reviewed by: kan
Threading library calls _pre before the fork, allowing the rtld to
lock itself to ensure that other threads of the process are out of
dynamic linker. _post releases the locks.
This allows the rtld to have consistent state in the child. Although
child may legitimately call only async-safe functions, the call may
need plt relocation resolution, and this requires working rtld.
Reported and debugging help by: rink
Reviewed by: kan, davidxu
MFC after: 1 month (anyway, not before 7.1 is out)
contexts as rtld's malloc is not thread safe and is only supposed to be
called with exclusive bind lock already held.
The originating PR submitted a patch on top of different pre-requisite
workaroud for unsafe dlopen calls, and the patch was midief slighlty to apply
to stock sources for the purpose of this commit. Running rtld malloc from
unlocked contexts is a bug on its own.
PR: 126950
Submited by: Oleg Dolgov
programs.
From the PR description:
The gcc runtime's _Unwind_Find_FDE function, invoked during exception
handling's stack unwinding, is not safe to execute from within multiple
threads. FreeBSD' s dl_iterate_phdr() however permits multiple threads
to pass through it though. The result is surprisingly reliable infinite
looping of one or more threads if they just happen to be unwinding at
the same time.
Introduce the new lock that is write locked around the dl_iterate_pdr,
thus providing required exclusion for the stack unwinders.
PR: threads/123062
Submitted by: Andy Newman <an at atrn org>
Reviewed by: kan
MFC after: 2 weeks
This code came from the merged mips2 and Juniper mips repositories.
Warner Losh, Randall Seager, Oleksandr Tymoshenko and Olivier Houchard
worked to merge, debug and integrate this code. This code may also
contain code derived from NetBSD.
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
main object list, its versioning information needs to be examined
separately.
This hopefully fixes problems that people running with SYMVER_ENABLED
are experiencing.
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
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)
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
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.
to be compatible with symbol versioning support as implemented by
GNU libc and documented by http://people.redhat.com/~drepper/symbol-versioning
and LSB 3.0.
Implement dlvsym() function to allow lookups for a specific version of
a given symbol.
means:
o Remove Elf64_Quarter,
o Redefine Elf64_Half to be 16-bit,
o Redefine Elf64_Word to be 32-bit,
o Add Elf64_Xword and Elf64_Sxword for 64-bit entities,
o Use Elf_Size in MI code to abstract the difference between
Elf32_Word and Elf64_Word.
o Add Elf_Ssize as the signed counterpart of Elf_Size.
MFC after: 2 weeks
we included the length of the path in the returned size but not the length
of the associated Dl_serpath structure. Without this fix, programs
attempting to allocate a structure to hold the search path information
would allocate too small of a buffer and rtld would overrun the buffer
while filling it via a subsequent RTLD_DI_SERINFO request.
Submitted by: "William K. Josephson" wkj at morphisms dot net
Reviewed by: jdp
MFC after: 2 weeks
are initialised to zero. When freeing TLS, don't attempt to free DTV
slots which were not used.
Pointed out by: Joerg Sonnenberger
X-MFC-After: After the branch, probably
LD_LIBMAP_DISABLE, LD_LIBRARY_PATH) are used, then make sure the
libraries being loaded aren't on a noexec-mounted filesystem.
This is a compromise position: I'm assuming that nobody will be silly
enough to set the noexec mount flag on part of the default library
path, in order to avoid adding extra overhead into the common case
(where those environment variables aren't used).
Discussed with: csjp, secteam
MFC after: 1 week
Another handy libmap patch. Lets you do stuff like this:
LD_LIBMAP="libpthread.so.1=libthr.so.1" mythreadedapp
If you already have a program-specific override in libmap.conf, note
that you must use a program-specific override in LD_LIBMAP:
LD_LIBMAP="[mythreadedapp],libpthread.so.1=libthr.so.1" mythreadedapp
PR: bin/74471
Submitted by: Dan Nelson <dnelson AT allantgroup.com>
MFC after: 2 weeks
(and that is for now being worked around by a binutils patch).
The rtld code tested &_DYNAMIC against 0 to see whether rtld itself
was built as PIC or not. While the sparc64 MD code did not rely
on the preset value of the GOT slot for _DYNAMIC any more due
to previous binutils changes, it still used to not be 0, so
that this check did work. The new binutils do however initialize
this slot with 0. As a consequence, rtld would not properly initialize
itself and crash.
Fix that by introducing a new macro, RTLD_IS_DYNAMIC, to take the role
of this test. For sparc64, it is implemented using the rtld_dynamic()
code that was already there. If an architecture does not provide its
own implementation, we default to the old check.
While being there, mark _DYNAMIC as a weak symbol in the sparc64
rtld_start.S. This is needed in the LDSCRIPT case, which is however
not currently supported for want of an actual ldscript.
Sanity checked with md5 on alpha, amd64, i386 and ia64.
that this provokes. "Wherever possible" means "In the kernel OR NOT
C++" (implying C).
There are places where (void *) pointers are not valid, such as for
function pointers, but in the special case of (void *)0, agreement
settles on it being OK.
Most of the fixes were NULL where an integer zero was needed; many
of the fixes were NULL where ascii <nul> ('\0') was needed, and a
few were just "other".
Tested on: i386 sparc64
(libmap available) and 1 for failure. Assign this return to the
global 'libmap_disable' variable in rtld.c.
This totally prevents any libmap functions from being called after
lm_init() if no config file is present.
Setting the LD_DUMP_REL_PRE or LD_DUMP_REL_POST environment variables
cause rtld-elf to output a table of all relocations.
This is useful for debugging.
implementation in case default one provided by rtld is
not suitable.
Consolidate various identical MD lock implementation into
a single file using appropriate machine/atomic.h.
Approved by: re (scottl)
DT_NEEDED links is not flexible enough for cases where dynamically
loaded modules form a dependency cycle.
This should fix an infinite recursion problem encountered by Yahoo.
Approved by: re (jhb)
This is an optional feature, disabled by default.
This will be useful to people testing the various POSIX threading
libraries under -CURRENT but can easily serve other needs.
Introduce a new unlink_object() function and call it in
unload_object() instead. Removing the object in unref_dag() is
too early, rtld calls _fini() function after that and shared
objects might fail resolve their own symbols.
Introdice RTLD_SELF special handle and properly process it within
dlsym() and dlinfo() functions.
The intention is to improve our compatibility with Solaris and
to make a Java port easier.
Partially submitted by: phantom
associated lists:
remove RTLD_GLOBAL objects from global objects list;
remove the parent object from dldags list of its children.
Previosly we were doing that only to the top-level object OF the DAG
being unloaded and all its dependencies were ignored, leading to
mysterious crashes later.
Submitted by: peter (partially)
before referencing object's DAG. This makes it possible for
C++ exceptions to work across shared libraries and brings
us closer to the search order used by Solaris/Linux.
Reviewed by: jdp
Approved by: obrien
MFC after: 1 month
even if there was no error occured (when trying to dlopen(3) object that
already linked into executable which does dlopen(3) call). This is more
proper fix for `ldd /usr/lib/libc.so' problem, because the new behaviour
conforms to documentation.
Remove workaround from ldd.c (rev.1.32).
PR: 35099
Submitted by: Nathan Hawkins <utsl@quic.net>
MFC after: 1 week
objects' reference counts. This function is called by the atexit
mechanism at program shutdown. I don't think the locking is necessary
here. It caused OpenOffice builds to hang more often than not.
Credit to Martin Blapp and Matt Dillon for helping to diagnose this
problem and for testing the fix.
o Set st_shndx for sym_zero to SHN_UNDEF instead of SHN_ABS.
This gives us something to reliably test against.
o For weak references to undefined sysmbols (as indicated by
having st_shndx equals SHN_UNDEF) in the context of OPDs,
the address of the OPD is to be zero, not the address of
the function it contains.
o For weak references to undefined symbols in all other cases
(only DIR64LSB at this time), the actual relocated value is
to be zero, not the value prior to relocating.
Roughly speaking, weak references to undefined symbols are no-ops.
Tested on: i386, ia64
relocation identifies the symbol to which we need to bind. This
solves a problem seen on ia64 where the symbol hash table does not
contain local symbols and thus resulted in unresolved symbols.
Tested on: alpha, i386, ia64
DT_INIT and DT_FINI tags pointed to fptr records. In 2.11.2, it points
to the actuall address of the function. On IA64 you cannot just take
an address of a function, store it in a function pointer variable and
call it.. the function pointers point to a fptr data block that has the
target gp and address in it. This is absolutely necessary for using
the in-tree binutils toolchain, but (unfortunately) will not work with
old shared libraries. Save your old ld-elf.so.1 if you want to use
old ones still. Do not mix-and-match.
This is a no-op change for i386 and alpha.
Reviewed by: dfr
particularly help programs which load many shared libraries with
a lot of relocations. Large C++ programs such as are found in KDE
are a prime example.
While relocating a shared object, maintain a vector of symbols
which have already been looked up, directly indexed by symbol
number. Typically, symbols which are referenced by a relocation
entry are referenced by many of them. This is the same optimization
I made to the a.out dynamic linker in 1995 (rtld.c revision 1.30).
Also, compare the first character of a sought-after symbol with its
symbol table entry before calling strcmp().
On a PII/400 these changes reduce the start-up time of a typical
KDE program from 833 msec (elapsed) to 370 msec.
MFC after: 5 days
The function's callers generate the error message when appropriate.
This eliminates the message ``Undefined symbol "__register_frame_info"''
which was bogusly returned by dlerror() in some cases.
function, thus allowing a debugger or other trace tool
to easily grab the addresses of the needed structures
off the stack.
This change is transparent to gdb, which locates the
link_map list and transfers it to debugger memory
for comparison purposes.
A sample program will be committed showing how this can
be used.
Reviewed by: John Polstra <jdp@FreeBSD.org>
Formerly the init functions were called in the opposite of the
order in which libraries were loaded, and libraries were loaded
according to a breadth-first traversal of the dependency graph.
That ordering came from SVR4.0, and it was easy to implement but
not always sensible.
Now we do a depth-first walk over the dependency graph and call
the init functions in an order such that each shared object's needed
objects are initialized before the shared object itself. At the
same time we build a list of finalization (fini) functions in the
opposite order, to guarantee correct C++ destructor ordering whenever
possible. (It may not be possible if dlopen and dlclose are used
in strange ways, but we come as close as one can come.)
The need for this renovation has become apparent as more programs
have started using multithreading. The multithreaded C library
libc_r requires initialization, whereas the standard libc does not.
Since virtually every other object depends on the C library, it is
important that it get initialized first.
and for all (I hope). Packages such as wine, JDK, and linuxthreads
should no longer have any problems with re-entering the dynamic
linker.
This commit replaces the locking used in the dynamic linker with a
new spinlock-based reader/writer lock implementation. Brian
Fundakowski Feldman <green> argued for this from the very beginning,
but it took me a long time to come around to his point of view.
Spinlocks are the only kinds of locks that work with all thread
packages. But on uniprocessor systems they can be inefficient,
because while a contender for the lock is spinning the holder of the
lock cannot make any progress toward releasing it. To alleviate
this disadvantage I have borrowed a trick from Sleepycat's Berkeley
DB implementation. When spinning for a lock, the requester does a
nanosleep() call for 1 usec. each time around the loop. This will
generally yield the CPU to other threads, allowing the lock holder
to finish its business and release the lock. I chose 1 usec. as the
minimum sleep which would with reasonable certainty not be rounded
down to 0.
The formerly machine-independent file "lockdflt.c" has been moved
into the architecture-specific subdirectories by repository copy.
It now contains the machine-dependent spinlocking code. For the
spinlocks I used the very nifty "simple, non-scalable reader-preference
lock" which I found at
<http://www.cs.rochester.edu/u/scott/synchronization/pseudocode/rw.html>
on all CPUs except the 80386 (the specific CPU model, not the
architecture). The 80386 CPU doesn't support the necessary "cmpxchg"
instruction, so on that CPU a simple exclusive test-and-set lock
is used instead. 80386 CPUs are detected at initialization time by
trying to execute "cmpxchg" and catching the resulting SIGILL
signal.
To reduce contention for the locks, I have revamped a couple of
key data structures, permitting all common operations to be done
under non-exclusive (reader) locking. The only operations that
require exclusive locking now are the rare intrusive operations
such as dlopen() and dlclose().
The dllockinit() interface is now deprecated. It still exists,
but only as a do-nothing stub. I plan to remove it as soon as is
reasonably possible. (From the very beginning it was clearly
labeled as experimental and subject to change.) As far as I know,
only the linuxthreads port uses dllockinit(). This interface turned
out to have several problems. As one example, when the dynamic
linker called a client-supplied locking function, that function
sometimes needed lazy binding, causing re-entry into the dynamic
linker and a big looping mess. And in any case, it turned out to be
too burdensome to require threads packages to register themselves
with the dynamic linker.
figure out which shared object(s) contain the the locking methods
and fully bind those objects as if they had been loaded with
LD_BIND_NOW=1. The goal is to keep the locking methods from
requiring any lazy binding. Otherwise infinite recursion occurs
in _rtld_bind.
This fixes the infinite recursion problem in the linuxthreads port.
init and fini functions. Now the code is very careful to hold no
locks when calling these functions. Thus the dynamic linker cannot
be re-entered with a lock already held.
Remove the tolerance for recursive locking that I added in revision
1.2 of dllockinit.c. Recursive locking shouldn't happen any more.
Mozilla and JDK users: I'd appreciate confirmation that things still
work right (or at least the same) with these changes.
functions to be used by the dynamic linker. This can be called by
threads packages at start-up time. I will add the call to libc_r
soon.
Also add a default locking method that is used up until dllockinit()
is called. The default method works by blocking SIGVTALRM, SIGPROF,
and SIGALRM in critical sections. It is based on the observation
that most user-space threads packages implement thread preemption
with one of these signals (usually SIGVTALRM).
The dynamic linker has never been reentrant, but it became less
reentrant in revision 1.34 of "src/libexec/rtld-elf/rtld.c".
Starting with that revision, multiple threads each doing lazy
binding could interfere with each other. The usual symptom was
that a symbol was falsely reported as undefined at start-up time.
It was rare but not unseen. This commit fixes it.
PT_INTERP program header entry, to ensure that gdb always finds
the right dynamic linker.
Use obj->relocbase to simplify a few calculations where appropriate.
loaded separately by dlopen that have global symbols with identical
names. Viewing each dlopened object as a DAG which is linked by its
DT_NEEDED entries in the dynamic table, the search order is as
follows:
* If the referencing object was linked with -Bsymbolic, search it
internally.
* Search all dlopened DAGs containing the referencing object.
* Search all objects loaded at program start up.
* Search all objects which were dlopened() using the RTLD_GLOBAL
flag (which is now supported too).
The search terminates as soon as a strong definition is found.
Lacking that, the first weak definition is used.
These rules match those of Solaris, as best I could determine them
from its vague manual pages and the results of experiments I performed.
PR: misc/12438
