The symbol was moved using scripts/move-symbol-to-libc.py.
It is necessary to arrange for a
__libpthread_version_placeholder@GLIBC_2.6 on some of the powerpc
targets.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
For some reason only dlopen failure caused dtv gaps to be reused.
It is possible that the intent was to never reuse modids for a
different module, but after dlopen failure all gaps are reused
not just the ones caused by the unfinished dlopened.
So the code has to handle reused modids already which seems to
work, however the data races at thread creation and tls access
(see bug 19329 and bug 27111) may be more severe if slots are
reused so this is scheduled after those fixes. I think fixing
the races are not simpler if reuse is disallowed and reuse has
other benefits, so set GL(dl_tls_dtv_gaps) whenever entries are
removed from the middle of the slotinfo list. The value does
not have to be correct: incorrect true value causes the next
modid query to do a slotinfo walk, incorrect false will leave
gaps and new entries are added at the end.
Fixes bug 27135.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Test concurrent dlopen and pthread_create when the loaded modules have
TLS. This triggers dl-tls assertion failures more reliably than the
nptl/tst-stack4 test.
The dlopened module has 100 DT_NEEDED dependencies with TLS, they were
reused from an existing TLS test. The number of created threads during
dlopen depends on filesystem speed and hardware, but at most 3 threads
are alive at a time to limit resource usage.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This is a follow up patch to the fix for bug 19329. This adds relaxed
MO atomics to accesses that were previously data races but are now
race conditions, and where relaxed MO is sufficient.
The race conditions all follow the pattern that the write is behind the
dlopen lock, but a read can happen concurrently (e.g. during tls access)
without holding the lock. For slotinfo entries the read value only
matters if it reads from a synchronized write in dlopen or dlclose,
otherwise the related dtv entry is not valid to access so it is fine
to leave it in an inconsistent state. The same applies for
GL(dl_tls_max_dtv_idx) and GL(dl_tls_generation), but there the
algorithm relies on the fact that the read of the last synchronized
write is an increasing value.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
DTV setup at thread creation (_dl_allocate_tls_init) is changed
to take the dlopen lock, GL(dl_load_lock). Avoiding data races
here without locks would require design changes: the map that is
accessed for static TLS initialization here may be concurrently
freed by dlclose. That use after free may be solved by only
locking around static TLS setup or by ensuring dlclose does not
free modules with static TLS, however currently every link map
with TLS has to be accessed at least to see if it needs static
TLS. And even if that's solved, still a lot of atomics would be
needed to synchronize DTV related globals without a lock. So fix
both bug 19329 and bug 27111 with a lock that prevents DTV setup
running concurrently with dlopen or dlclose.
_dl_update_slotinfo at TLS access still does not use any locks
so CONCURRENCY NOTES are added to explain the synchronization.
The early exit from the slotinfo walk when max_modid is reached
is not strictly necessary, but does not hurt either.
An incorrect acquire load was removed from _dl_resize_dtv: it
did not synchronize with any release store or fence and
synchronization is now handled separately at thread creation
and TLS access time.
There are still a number of racy read accesses to globals that
will be changed to relaxed MO atomics in a followup patch. This
should not introduce regressions compared to existing behaviour
and avoid cluttering the main part of the fix.
Not all TLS access related data races got fixed here: there are
additional races at lazy tlsdesc relocations see bug 27137.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
The symbols pthread_clockjoin_np, pthread_join, pthread_timedjoin_np,
pthread_tryjoin_np, thrd_join were moved using
scripts/move-symbol-to-libc.py.
Moving the symbols at the same time avoids the need for temporary
exports.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This replaces the FREE_P macro with the __nptl_stack_in_use inline
function. stack_list_del is renamed to __nptl_stack_list_del,
stack_list_add to __nptl_stack_list_add, __deallocate_stack to
__nptl_deallocate_stack, free_stacks to __nptl_free_stacks.
It is convenient to move __libpthread_freeres into libc at the
same time. This removes the temporary __default_pthread_attr_freeres
export and restores full freeres coverage for __default_pthread_attr.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
The symbol was moved using scripts/move-symbol-to-libc.py.
The export of __default_pthread_attr_freeres is temporary. There
is a minor regression in freeres coverage because in the dynamic case,
__default_pthread_attr_freeres is no longer called if libpthread is
not linked in.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This removes the DEBUGGING_P macro and the __pthread_debug variable.
The __find_in_stack_list function is now unused and deleted as well.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
The nptl version is used as default, since now with symbol always
present the single-thread optimization is tricky.
Hurd is not change, it is used it own lock scheme (which call
_cthreads_funlockfile).
Checked on x86_64-linux-gnu.
The nptl version is used as default, since now with symbol always
present the single-thread optimization is tricky.
Hurd is not change, it is used it own lock scheme (which call
_cthreads_ftrylockfile).
Checked on x86_64-linux-gnu.
The nptl version is used as default, since now with symbol always
present the single-thread optimization is tricky.
Hurd is not change, it is used it own lock scheme (which call
_cthreads_flockfile).
Checked on x86_64-linux-gnu.
Linux 5.12 adds the constants PTRACE_SYSEMU and
PTRACE_SYSEMU_SINGLESTEP for s390. Add these to glibc.
Tested with build-many-glibcs.py for s390-linux-gnu and
s390x-linux-gnu.
These workload traces cover the whole "long double" range.
This patch was prepared with the help of Adhemerval Zanella.
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
The GLIBC_PRIVATE exports for these symbols are expected to be
temporary.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
stack_list_del overwrites the in-flight stack variable.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
All the stack lists are now in _rtld_global, so it is possible
to change stack permissions directly from there, instead of
calling into libpthread to do the change.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
Only ia64 needs the page mask, and it is straightforward
to compute the value within the function itself.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
Permissions of the cached stacks may have to be updated if an object
is loaded that requires executable stacks, so the dynamic loader
needs to know about these cached stacks.
The move of in_flight_stack and stack_cache_actsize is a requirement for
merging __reclaim_stacks into the fork implementation in libc.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
This is an early variant of __tls_init_tp, primarily for initializing
thread-related elements of _rtld_global/GL.
Some existing initialization code not needed for NPTL is moved into
the generic version of this function.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
It is no longer needed after the SINGLE_THREADED_P consolidation.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
Always use __libc_multiple_threads if beneficial, and do not assume
the the dynamic loader is single-threaded. This assumption could
become incorrect by accident once more code is moved from libpthread
into it. The previous commit introducing the
NO_SYSCALL_CANCEL_CHECKING macro enables this change.
Do not hint to the compiler that multi-threaded programs are unlikely
(which is not quite true anymore).
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
Historically, SINGLE_THREAD_P is defined to 1 in the dynamic loader.
This has the side effect of disabling cancellation points. In order
to enable future use of SINGLE_THREAD_P for single-thread
optimizations in the dynamic loader (which becomes important once
more code is moved from libpthread), introduce a new
NO_SYSCALL_CANCEL_CHECKING macro which is always 1 for IS_IN (rtld),
indepdently of the actual SINGLE_THREAD_P value.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
This allows the elimination of the __libc_multiple_threads_ptr
variable in libpthread and its initialization procedure.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
If libpthread is included in libc, it is not necessary to delay
initialization of the lock/unlock function pointers until libpthread
is loaded. This eliminates two unprotected function pointers
from _rtld_global and removes some initialization code from
libpthread.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
This generates new macros of this from:
They are useful for symbol lookups using _dl_lookup_direct.
Tested-by: Carlos O'Donell <carlos@redhat.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
No bug.
This commit adds a new implementation for EVEX memchr that is not safe
for RTM because it uses vzeroupper. The benefit is that by using
ymm0-ymm15 it can use vpcmpeq and vpternlogd in the 4x loop which is
faster than the RTM safe version which cannot use vpcmpeq because
there is no EVEX encoding for the instruction. All parts of the
implementation aside from the 4x loop are the same for the two
versions and the optimization is only relevant for large sizes.
Tigerlake:
size , algn , Pos , Cur T , New T , Win , Dif
512 , 6 , 192 , 9.2 , 9.04 , no-RTM , 0.16
512 , 7 , 224 , 9.19 , 8.98 , no-RTM , 0.21
2048 , 0 , 256 , 10.74 , 10.54 , no-RTM , 0.2
2048 , 0 , 512 , 14.81 , 14.87 , RTM , 0.06
2048 , 0 , 1024 , 22.97 , 22.57 , no-RTM , 0.4
2048 , 0 , 2048 , 37.49 , 34.51 , no-RTM , 2.98 <--
Icelake:
size , algn , Pos , Cur T , New T , Win , Dif
512 , 6 , 192 , 7.6 , 7.3 , no-RTM , 0.3
512 , 7 , 224 , 7.63 , 7.27 , no-RTM , 0.36
2048 , 0 , 256 , 8.48 , 8.38 , no-RTM , 0.1
2048 , 0 , 512 , 11.57 , 11.42 , no-RTM , 0.15
2048 , 0 , 1024 , 17.92 , 17.38 , no-RTM , 0.54
2048 , 0 , 2048 , 30.37 , 27.34 , no-RTM , 3.03 <--
test-memchr, test-wmemchr, and test-rawmemchr are all passing.
Signed-off-by: Noah Goldstein <goldstein.w.n@gmail.com>
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
An unknown vector operation occurred in commit 2a76821c30. Fixed it
by using "ymm{k1}{z}" but not "ymm {k1} {z}".
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
The hwcap2 check for the aforementioned functions should check for
both PPC_FEATURE2_ARCH_3_1 and PPC_FEATURE2_HAS_ISEL but was
mistakenly checking for any one of them, enabling isa 3.1 version of
the functions in incompatible processors, like POWER8.
Reviewed-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
Big win in binary size and avoids duplicating the logic in multiple
places.
On x86_64, dropped from 1883206 to 1881790, a 1416 byte decrease.
Also changed logic to track if ttyname_buf has been allocated by
checking if it's NULL instead of tracking buflen as an additional
variable.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Simplifies the logic and makes intent clearer, while at the same time
decreasing binary size.
On x86_64, dropped from 1883270 to 1883206, a 64 byte decrease.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
In a default build for x86_64, size decreased by 24 bytes:
1883294 to 1883270.
Aditionally, avoids repeating the number printing logic in multiple
places.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
