about summary refs log tree commit diff
path: root/src/thread/i386
Commit message (Collapse)AuthorAgeFilesLines
* install dynamic tls synchronously at dlopen, streamline accessRich Felker2019-02-181-8/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | previously, dynamic loading of new libraries with thread-local storage allocated the storage needed for all existing threads at load-time, precluding late failure that can't be handled, but left installation in existing threads to take place lazily on first access. this imposed an additional memory access and branch on every dynamic tls access, and imposed a requirement, which was not actually met, that the dynamic tlsdesc asm functions preserve all call-clobbered registers before calling C code to to install new dynamic tls on first access. the x86[_64] versions of this code wrongly omitted saving and restoring of fpu/vector registers, assuming the compiler would not generate anything using them in the called C code. the arm and aarch64 versions saved known existing registers, but failed to be future-proof against expansion of the register file. now that we track live threads in a list, it's possible to install the new dynamic tls for each thread at dlopen time. for the most part, synchronization is not needed, because if a thread has not synchronized with completion of the dlopen, there is no way it can meaningfully request access to a slot past the end of the old dtv, which remains valid for accessing slots which already existed. however, it is necessary to ensure that, if a thread sees its new dtv pointer, it sees correct pointers in each of the slots that existed prior to the dlopen. my understanding is that, on most real-world coherency architectures including all the ones we presently support, a built-in consume order guarantees this; however, don't rely on that. instead, the SYS_membarrier syscall is used to ensure that all threads see the stores to the slots of their new dtv prior to the installation of the new dtv. if it is not supported, the same is implemented in userspace via signals, using the same mechanism as __synccall. the __tls_get_addr function, variants, and dynamic tlsdesc asm functions are all updated to remove the fallback paths for claiming new dynamic tls, and are now all branch-free.
* make arch __set_thread_area backends hiddenRich Felker2018-09-121-0/+1
| | | | | | | | this is not a public interface, and does not even necessarily match the syscall on all archs that have a syscall by that name. on archs where it's implemented in C, no action on the source file is needed; the hidden declaration in pthread_arch.h suffices.
* make arch __clone backends hiddenRich Felker2018-09-121-0/+1
| | | | | these are not a public interface and are not intended to be callable from anywhere but the public clone function or other places in libc.
* in i386 __set_thread_area, don't assume %gs register is initially zeroRich Felker2015-05-161-4/+9
| | | | | | | | | | | | | | | | | | | | | | | commit f630df09b1fd954eda16e2f779da0b5ecc9d80d3 added logic to handle the case where __set_thread_area is called more than once by reusing the GDT slot already in the %gs register, and only setting up a new GDT slot when %gs is zero. this created a hidden assumption that %gs is zero when a new process image starts, which is true in practice on Linux, but does not seem to be documented ABI, and fails to hold under qemu app-level emulation. while it would in theory be possible to zero %gs in the entry point code, this code is shared between static and dynamic binaries, and dynamic binaries must not clobber the value of %gs already setup by the dynamic linker. the alternative solution implemented in this commit simply uses global data to store the GDT index that's selected. __set_thread_area should only be called in the initial thread anyway (subsequent threads get their thread pointer setup by __clone), but even if it were called by another thread, it would simply read and write back the same GDT index that was already assigned to the initial thread, and thus (in the x86 memory model) there is no data race.
* use hidden __tls_get_new for tls/tlsdesc lookup fallback casesRich Felker2015-04-141-1/+3
| | | | | | | | | | | | | | | previously, the dynamic tlsdesc lookup functions and the i386 special-ABI ___tls_get_addr (3 underscores) function called __tls_get_addr when the slot they wanted was not already setup; __tls_get_addr would then in turn also see that it's not setup and call __tls_get_new. calling __tls_get_new directly is both more efficient and avoids the issue of calling a non-hidden (public API/ABI) function from asm. for the special i386 function, a weak reference to __tls_get_new is used since this function is not defined when static linking (the code path that needs it is unreachable in static-linked programs).
* consistently use hidden visibility for cancellable syscall internalsRich Felker2015-04-141-3/+8
| | | | | | | | | | in a few places, non-hidden symbols were referenced from asm in ways that assumed ld-time binding. while these is no semantic reason these symbols need to be hidden, fixing the references without making them hidden was going to be ugly, and hidden reduces some bloat anyway. in the asm files, .global/.hidden directives have been moved to the top to unclutter the actual code.
* allow i386 __set_thread_area to be called more than onceRich Felker2015-04-131-1/+5
| | | | | | | | previously a new GDT slot was requested, even if one had already been obtained by a previous call. instead extract the old slot number from GS and reuse it if it was already set. the formula (GS-3)/8 for the slot number automatically yields -1 (request for new slot) if GS is zero (unset).
* prepare cancellation syscall asm for possibility of __cancel returningRich Felker2015-02-201-1/+8
|
* optimize i386 ___tls_get_addr asmRich Felker2014-06-191-1/+8
|
* add thread-pointer support for pre-2.6 kernels on i386Rich Felker2014-06-101-4/+18
| | | | | | | | | | | | | | | | | | | such kernels cannot support threads, but the thread pointer is also important for other purposes, most notably stack protector. without a valid thread pointer, all code compiled with stack protector will crash. the same applies to any use of thread-local storage by applications or libraries. the concept of this patch is to fall back to using the modify_ldt syscall, which has been around since linux 1.0, to setup the gs segment register. since the kernel does not have a way to automatically assign ldt entries, use of slot zero is hard-coded. if this fallback path is used, __set_thread_area returns a positive value (rather than the usual zero for success, or negative for error) indicating to the caller that the thread pointer was successfully set, but only for the main thread, and that thread creation will not work properly. the code in __init_tp has been changed accordingly to record this result for later use by pthread_create.
* clone: make clone a wrapper around __cloneBobby Bingham2014-02-091-3/+0
| | | | | | | | | | | | The architecture-specific assembly versions of clone did not set errno on failure, which is inconsistent with glibc. __clone still returns the error via its return value, and clone is now a wrapper that sets errno as needed. The public clone has also been moved to src/linux, as it's not directly related to the pthreads API. __clone is called by pthread_create, which does not report errors via errno. Though not strictly necessary, it's nice to avoid clobbering errno here.
* beginnings of full TLS support in shared librariesRich Felker2012-10-041-0/+8
| | | | | | this code will not work yet because the necessary relocations are not supported, and cannot be supported without some internal changes to how relocation processing works (coming soon).
* fix issue with longjmp out of signal handlers and cancellationRich Felker2012-05-231-21/+15
| | | | | | | | | | | | | | | | | | | stale state information indicating that a thread was possibly blocked at a cancellation point could get left behind if longjmp was used to exit a signal handler that interrupted a cancellation point. to fix the issue, we throw away the state information entirely and simply compare the saved instruction pointer to a range of code addresses in the __syscall_cp_asm function. all the ugly PIC work (which becomes minimal anyway with this approach) is defered to cancellation time instead of happening at every syscall, which should improve performance too. this commit also fixes cancellation on arm, which was mildly broken (race condition, not checking cancellation flag once inside the cancellation point zone). apparently i forgot to implement that. the new arm code is untested, but appears correct; i'll test and fix it later if there are problems.
* replace bad cancellation cleanup abi with a sane oneRich Felker2012-02-091-34/+0
| | | | | | | | | | | | | | | | | | | | | | the old abi was intended to duplicate glibc's abi at the expense of being ugly and slow, but it turns out glib was not even using that abi except on non-gcc-compatible compilers (which it doesn't even support) and was instead using an exceptions-in-c/unwind-based approach whose abi we could not duplicate anyway without nasty dwarf2/unwind integration. the new abi is copied from a very old glibc abi, which seems to still be supported/present in current glibc. it avoids all unwinding, whether by sjlj or exceptions, and merely maintains a linked list of cleanup functions to be called from the context of pthread_exit. i've made some care to ensure that longjmp out of a cleanup function should work, even though it is not required to. this change breaks abi compatibility with programs which were using pthread cancellation, which is unfortunate, but that's why i'm making the change now rather than later. considering that most pthread features have not been usable until recently anyway, i don't see it as a major issue at this point.
* overhaul clone syscall wrappingRich Felker2011-09-182-34/+60
| | | | | | | | | | | | | | | | | | | | several things are changed. first, i have removed the old __uniclone function signature and replaced it with the "standard" linux __clone/clone signature. this was necessary to expose clone to applications anyway, and it makes it easier to port __clone to new archs, since it's now testable independently of pthread_create. secondly, i have removed all references to the ugly ldt descriptor structure (i386 only) from the c code and pthread structure. in places where it is needed, it is now created on the stack just when it's needed, in assembly code. thus, the i386 __clone function takes the desired thread pointer as its argument, rather than an ldt descriptor pointer, just like on all other sane archs. this should not affect applications since there is really no way an application can use clone with threads/tls in a way that doesn't horribly conflict with and clobber the underlying implementation's use. applications are expected to use clone only for creating actual processes, possibly with new namespace features and whatnot.
* further debloat cancellation handlersRich Felker2011-08-031-3/+13
| | | | | | | cleanup push and pop are also no-ops if pthread_exit is not reachable. this can make a big difference for library code which needs to protect itself against cancellation, but which is unlikely to actually be used in programs with threads/cancellation.
* fix static linking dependency bloat with cancellationRich Felker2011-08-032-8/+8
| | | | | | | previously, pthread_cleanup_push/pop were pulling in all of pthread_create due to dependency on the __pthread_unwind_next function. this was not needed, as cancellation cleanup handlers can never be called unless pthread_exit or pthread_cancel is reachable.
* restore use of .type in asm, but use modern @function (vs %function)Rich Felker2011-06-146-0/+7
| | | | | | | | this seems to be necessary to make the linker accept the functions in a shared library (perhaps to generate PLT entries?) strictly speaking libc-internal asm should not need it. i might clean that up later.
* remove all .size and .type directives for functions from the asmRich Felker2011-06-136-11/+0
| | | | | these are useless and have caused problems for users trying to build with non-gnu tools like tcc's assembler.
* fix bugs in cancellable syscall asmRich Felker2011-04-171-5/+4
| | | | | | | | | | | | | | | | x86_64 was just plain wrong in the cancel-flag-already-set path, and crashing. the more subtle error was not clearing the saved stack pointer before returning to c code. this could result in the signal handler misidentifying c code as the pre-syscall part of the asm, and acting on cancellation at the wrong time, and thus resource leak race conditions. also, now __cancel (in the c code) is responsible for clearing the saved sp in the already-cancelled branch. this means we have to use call rather than jmp to ensure the stack pointer in the c will never match what the asm saved.
* fix some minor issues in cancellation handling patchRich Felker2011-04-171-1/+1
| | | | | signals were wrongly left masked, and cancellability state was not switched to disabled, during the execution of cleanup handlers.
* overhaul pthread cancellationRich Felker2011-04-171-0/+36
| | | | | | | | | | | | | | | | | | | | | | this patch improves the correctness, simplicity, and size of cancellation-related code. modulo any small errors, it should now be completely conformant, safe, and resource-leak free. the notion of entering and exiting cancellation-point context has been completely eliminated and replaced with alternative syscall assembly code for cancellable syscalls. the assembly is responsible for setting up execution context information (stack pointer and address of the syscall instruction) which the cancellation signal handler can use to determine whether the interrupted code was in a cancellable state. these changes eliminate race conditions in the previous generation of cancellation handling code (whereby a cancellation request received just prior to the syscall would not be processed, leaving the syscall to block, potentially indefinitely), and remedy an issue where non-cancellable syscalls made from signal handlers became cancellable if the signal handler interrupted a cancellation point. x86_64 asm is untested and may need a second try to get it right.
* match glibc/lsb cancellation abi on i386Rich Felker2011-03-252-0/+24
| | | | | | | | glibc made the ridiculous choice to use pass-by-register calling convention for these functions, which is impossible to duplicate directly on non-gcc compilers. instead, we use ugly asm to wrap and convert the calling convention. presumably this works with every compiler anyone could potentially want to use.
* race condition fix: block all signals before decrementing thread countRich Felker2011-02-191-9/+0
| | | | | | | | | | | | the existence of a (kernelspace) thread must never have observable effects after the thread count is decremented. if signals are not blocked, it could end up handling the signal for rsyscall and contributing towards the count of threads which have changed ids, causing a thread to be missed. this could lead to one thread retaining unwanted privilege level. this change may also address other subtle race conditions in application code that uses signals.
* finish unifying thread register handling in preparation for portingRich Felker2011-02-151-0/+22
|
* begin unifying clone/thread management interface in preparation for portingRich Felker2011-02-151-25/+17
|
* initial check-in, version 0.5.0 v0.5.0Rich Felker2011-02-122-0/+57