Linux: Use rseq in sched_getcpu if available (v9)
Subsystems (3)
A message that is not in the archive (sent off-list or not yet received): [email protected]
PATCH 1/3 glibc: Perform rseq registration at C startup and thread creation (v20)
Register rseq TLS for each thread (including main), and unregister for each thread (excluding main). "rseq" stands for Restartable Sequences.
See the rseq(2) man page proposed here:
https://lkml.org/lkml/2018/9/19/647
Those are based on glibc master branch commit 415d0b0b3f16. The rseq system call was merged into Linux 4.18.
---
Changes since v1:
- Move __rseq_refcount to an extra field at the end of __rseq_abi to eliminate one symbol.
All libraries/programs which try to register rseq (glibc, early-adopter applications, early-adopter libraries) should use the rseq refcount. It becomes part of the ABI within a user-space process, but it's not part of the ABI shared with the kernel per se.
- Restructure how this code is organized so glibc keeps building on non-Linux targets.
- Use non-weak symbol for __rseq_abi.
- Move rseq registration/unregistration implementation into its own nptl/rseq.c compile unit.
- Move __rseq_abi symbol under GLIBC_2.29.
Changes since v2:
- Move __rseq_refcount to its own symbol, which is less ugly than trying to play tricks with the rseq uapi.
- Move __rseq_abi from nptl to csu (C start up), so it can be used across glibc, including memory allocator and sched_getcpu(). The __rseq_refcount symbol is kept in nptl, because there is no reason to use it elsewhere in glibc.
Changes since v3:
- Set __rseq_refcount TLS to 1 on register/set to 0 on unregister because glibc is the first/last user.
- Unconditionally register/unregister rseq at thread start/exit, because glibc is the first/last user.
- Add missing abilist items.
- Rebase on glibc master commit a502c5294.
- Add NEWS entry.
Changes since v4:
- Do not use "weak" symbols for __rseq_abi and __rseq_refcount. Based on "System V Application Binary Interface", weak only affects the link editor, not the dynamic linker.
- Install a new sys/rseq.h system header on Linux, which contains the RSEQ_SIG definition, __rseq_abi declaration and __rseq_refcount declaration. Move those definition/declarations from rseq-internal.h to the installed sys/rseq.h header.
- Considering that rseq is only available on Linux, move csu/rseq.c to sysdeps/unix/sysv/linux/rseq-sym.c.
- Move __rseq_refcount from nptl/rseq.c to sysdeps/unix/sysv/linux/rseq-sym.c, so it is only defined on Linux.
- Move both ABI definitions for __rseq_abi and __rseq_refcount to sysdeps/unix/sysv/linux/Versions, so they only appear on Linux.
- Document __rseq_abi and __rseq_refcount volatile.
- Document the RSEQ_SIG signature define.
- Move registration functions from rseq.c to rseq-internal.h static inline functions. Introduce empty stubs in misc/rseq-internal.h, which can be overridden by architecture code in sysdeps/unix/sysv/linux/rseq-internal.h.
- Rename __rseq_register_current_thread and __rseq_unregister_current_thread to rseq_register_current_thread and rseq_unregister_current_thread, now that those are only visible as internal static inline functions.
- Invoke rseq_register_current_thread() from libc-start.c LIBC_START_MAIN rather than nptl init, so applications not linked against libpthread.so have rseq registered for their main() thread. Note that it is invoked separately for SHARED and !SHARED builds.
Changes since v5:
- Replace __rseq_refcount by __rseq_lib_abi, which contains two
uint32_t: register_state and refcount. The "register_state" field allows inhibiting rseq registration from signal handlers nested on top of glibc registration and occuring after rseq unregistration by glibc.
- Introduce enum rseq_register_state, which contains the states allowed for the struct rseq_lib_abi register_state field.
Changes since v6:
- Introduce bits/rseq.h to define RSEQ_SIG for each architecture. The generic bits/rseq.h does not define RSEQ_SIG, meaning that each architecture implementing rseq needs to implement bits/rseq.h.
- Rename enum item RSEQ_REGISTER_NESTED to RSEQ_REGISTER_ONGOING.
- Port to glibc-2.29.
Changes since v7:
- Remove __rseq_lib_abi symbol, including refcount and register_state fields.
- Remove reference counting and nested signals handling from registration/unregistration functions.
- Introduce new __rseq_handled exported symbol, which is set to 1 by glibc on C startup when it handles restartable sequences. This allows glibc to coexist with early adopter libraries and applications wishing to register restartable sequences when it is not handled by glibc.
- Introduce rseq_init (), which sets __rseq_handled to 1 from C startup.
- Update NEWS entry.
- Update comments at the beginning of new files.
- Registration depends on both __NR_rseq and RSEQ_SIG.
- Remove ARM, powerpc, MIPS RSEQ_SIG until we agree with maintainers on the signature choice.
- Update x86, s390 RSEQ_SIG based on discussion with arch maintainers.
- Remove rseq-internal.h from headers list of misc/Makefile, so it it not installed by make install.
Changes since v8:
- Introduce RSEQ_SIG_CODE and RSEQ_SIG_DATA on aarch64 to handle compiling with -mbig-endian.
Changes since v9:
- Update Changelog.
- Remove unneeded new file comment header newlines.
Changes since v10:
- Remove volatile from __rseq_abi declaration.
- Document that __rseq_handled is about library managing rseq registration, independently of whether rseq is available or not.
- Move __rseq_handled symbol to ld.so, initialize this symbol within the dynamic linker initialization for both shared (rtld.c) and static (dl-support.c) builds.
- Only register the rseq TLS on initialization once in multiple-libc scenarios. Use rtld_active () for this purpose.
- In the static libc case, register the rseq TLS after LD_PRELOAD constructors are run, so it matches the order of this initialization vs LD_PRELOAD contructors execution for the shared libc.
- Agreed on signature choice with powerpc and MIPS maintainers, re-adding those signatures,
- The main architecture still left out signature-wise is ARM32.
Changes since v11:
- Rebase on glibc 2.30.
- Re-introduce ARM RSEQ_SIG following feedback from Will Deacon.
Changes since v12:
- Remove __rseq_handled,
- Rely on OS implicit rseq unregistration on thread teardown,
- Register main thread in __libc_early_init ().
- Add Restartable Sequences entry to threads manual.
Changes since v13:
- Update following be/le abilist split for arm, microblaze, and sh.
- Update manual to add the __rseq_abi variable and RSEQ_SIG macro to generate manual index entries, and add missing "Restartable Sequences" menu entry to the threads chapter.
Changes since v14:
- Update copyright range to include 2020.
- Introduce __ASSUME_RSEQ defined for --enable-kernel=4.18.0 and higher.
- Use ifdef __ASSUME_RSEQ rather than ifdef __NR_rseq to discover rseq availability. This is necessary now that the system call numbers are integrated within glibc.
Changes since v15:
- Remove __ASSUME_RSEQ from kernel features.
- rseq internal: remove assume rseq
- remove assume rseq and struct rseq def from sysdeps/unix/sysv/linux/rseq-sym.c
- sys/rseq.h: detect rseq header, implement fallback
- sysdeps/unix/sysv/linux/sys/rseq.h include cdefs.h, add _Static_assert to validate struct rseq and struct rseq_cs alignment.
- sys/rseq.h: document that posix_memalign should be used rather than malloc if allocating struct rseq or struct rseq_cs on the heap. This is required to guarantee 32-byte alignement.
Changes since v16:
- Move rseq NEWS entry under 2.32.
- Move new __rseq_abi symbol to GLIBC_2.32.
Changes since v17:
- Change copyright year to 2020.
- Refer to GNU C Library manual rather than rseq manpage in NEWS.
- Use "initial" parameter from __libc_early_init ().
- Manual: rseq is Linux rather than GNU std.
- Remove rseq_unregister_current_thread () (unused).
- rseq_register_current_thread () returns void.
- Coding style fixes.
- sys/rseq.h: use "32" for alignment.
- Change http:// for https:// in comments.
- Add const struct rseq_cs * field to rseq_cs union.
Changes since v18:
- NEWS update.
- Manual update.
- Move misc/rseq-internal.h to sysdeps/generic/rseq-internal.h.
- Fix coding style in sysdeps/unix/sysv/linux/sys/rseq.h.
- Abort libc if __rseq_abi is already initialized in rseq_register_current_thread (), which is unexpected.
- Abort libc in rseq_register_current_thread () on all errno except EPERM (seccomp), EACCES (seccomp), and ENOSYS (not implemented).
Changes since v19:
- Take care of feedback from Florian.
- Update manual to remove uptr field, which is removed from sys/rseq.h.
- Rebase on current master branch.
--- NEWS | 10 + elf/libc_early_init.c | 4 + manual/threads.texi | 63 +++++ nptl/pthread_create.c | 13 + sysdeps/generic/rseq-internal.h | 26 ++ sysdeps/unix/sysv/linux/Makefile | 5 +- sysdeps/unix/sysv/linux/Versions | 1 + sysdeps/unix/sysv/linux/aarch64/bits/rseq.h | 43 ++++ sysdeps/unix/sysv/linux/aarch64/libc.abilist | 1 + sysdeps/unix/sysv/linux/alpha/libc.abilist | 1 + sysdeps/unix/sysv/linux/arm/be/libc.abilist | 1 + sysdeps/unix/sysv/linux/arm/bits/rseq.h | 83 +++++++ sysdeps/unix/sysv/linux/arm/le/libc.abilist | 1 + sysdeps/unix/sysv/linux/bits/rseq.h | 29 +++ sysdeps/unix/sysv/linux/csky/libc.abilist | 1 + sysdeps/unix/sysv/linux/hppa/libc.abilist | 1 + sysdeps/unix/sysv/linux/i386/libc.abilist | 1 + sysdeps/unix/sysv/linux/ia64/libc.abilist | 1 + .../sysv/linux/m68k/coldfire/libc.abilist | 1 + .../unix/sysv/linux/m68k/m680x0/libc.abilist | 1 + .../sysv/linux/microblaze/be/libc.abilist | 1 + .../sysv/linux/microblaze/le/libc.abilist | 1 + sysdeps/unix/sysv/linux/mips/bits/rseq.h | 62 +++++ .../sysv/linux/mips/mips32/fpu/libc.abilist | 1 + .../sysv/linux/mips/mips32/nofpu/libc.abilist | 1 + .../sysv/linux/mips/mips64/n32/libc.abilist | 1 + .../sysv/linux/mips/mips64/n64/libc.abilist | 1 + sysdeps/unix/sysv/linux/nios2/libc.abilist | 1 + sysdeps/unix/sysv/linux/powerpc/bits/rseq.h | 37 +++ .../linux/powerpc/powerpc32/fpu/libc.abilist | 1 + .../powerpc/powerpc32/nofpu/libc.abilist | 1 + .../linux/powerpc/powerpc64/be/libc.abilist | 1 + .../linux/powerpc/powerpc64/le/libc.abilist | 1 + .../unix/sysv/linux/riscv/rv64/libc.abilist | 1 + sysdeps/unix/sysv/linux/rseq-internal.h | 73 ++++++ sysdeps/unix/sysv/linux/rseq-sym.c | 26 ++ sysdeps/unix/sysv/linux/s390/bits/rseq.h | 37 +++ .../unix/sysv/linux/s390/s390-32/libc.abilist | 1 + .../unix/sysv/linux/s390/s390-64/libc.abilist | 1 + sysdeps/unix/sysv/linux/sh/be/libc.abilist | 1 + sysdeps/unix/sysv/linux/sh/le/libc.abilist | 1 + .../sysv/linux/sparc/sparc32/libc.abilist | 1 + .../sysv/linux/sparc/sparc64/libc.abilist | 1 + sysdeps/unix/sysv/linux/sys/rseq.h | 228 ++++++++++++++++++ sysdeps/unix/sysv/linux/x86/bits/rseq.h | 30 +++ .../unix/sysv/linux/x86_64/64/libc.abilist | 1 + .../unix/sysv/linux/x86_64/x32/libc.abilist | 1 + 47 files changed, 798 insertions(+), 2 deletions(-) create mode 100644 sysdeps/generic/rseq-internal.h create mode 100644 sysdeps/unix/sysv/linux/aarch64/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/arm/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/mips/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/powerpc/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/rseq-internal.h create mode 100644 sysdeps/unix/sysv/linux/rseq-sym.c create mode 100644 sysdeps/unix/sysv/linux/s390/bits/rseq.h create mode 100644 sysdeps/unix/sysv/linux/sys/rseq.h create mode 100644 sysdeps/unix/sysv/linux/x86/bits/rseq.h
NEWS +10 -0
@@ -27,6 +27,16 @@ Major new features: several APIs have been annotated with GCC 'access' attribute. This should help GCC 10 issue better warnings. +* Support for automatically registering threads with the Linux rseq + system call has been added. This system call is implemented starting + from Linux 4.18. The Restartable Sequences ABI accelerates user-space + operations on per-cpu data. It allows user-space to perform updates + on per-cpu data without requiring heavy-weight atomic operations. + Automatically registering threads allows all libraries, including libc, + to make immediate use of the rseq support by using the documented ABI. + The GNU C Library manual has details on integration of Restartable + Sequences. + Deprecated and removed features, and other changes affecting compatibility: * The deprecated <sys/sysctl.h> header and the sysctl function have been
elf/libc_early_init.c +4 -0
@@ -18,10 +18,14 @@#include <ctype.h> #include <libc-early-init.h> +#include <rseq-internal.h> void __libc_early_init (_Bool initial) {/* Initialize ctype data. */ __ctype_init (); +/* Register rseq ABI to the kernel for the main program's libc. */ +if (initial) +rseq_register_current_thread (); }
manual/threads.texi +63 -0
75 lines not on this page · raw message
nptl/pthread_create.c +13 -0
@@ -33,6 +33,7 @@#include <default-sched.h> #include <futex-internal.h> #include <tls-setup.h> +#include <rseq-internal.h> #include "libioP.h" #include <shlib-compat.h>
@@ -384,6 +385,9 @@ START_THREAD_DEFN/* Initialize pointers to locale data. */ __ctype_init (); +/* Register rseq TLS to the kernel. */ +rseq_register_current_thread (); +#ifndef __ASSUME_SET_ROBUST_LIST if (__set_robust_list_avail >= 0)#endif
@@ -579,6 +583,15 @@ START_THREAD_DEFNprocess is really dead since 'clone' got passed the CLONE_CHILD_CLEARTID flag. The 'tid' field in the TCB will be set to zero. +rseq TLS is still registered at this point. Rely on implicit +unregistration performed by the kernel on thread teardown. This is not a +problem because the rseq TLS lives on the stack, and the stack outlives +the thread. If TCB allocation is ever changed, additional steps may be +required, such as performing explicit rseq unregistration before +reclaiming the rseq TLS area memory. It is NOT sufficient to block +signals because the kernel may write to the rseq area even without +signals. +The exit code is zero since in case all threads exit by calling 'pthread_exit' the exit status must be 0 (zero). */ __exit_thread ();
sysdeps/generic/rseq-internal.h new file +26 -0
@@ -0,0 +1,26 @@ +/* Restartable Sequences internal API. Stub version. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#ifndef RSEQ_INTERNAL_H +#define RSEQ_INTERNAL_H + +static inline void +rseq_register_current_thread (void ) +{ +} + +#endif /* rseq-internal.h */
sysdeps/unix/sysv/linux/Makefile +3 -2
@@ -41,7 +41,7 @@ update-syscall-lists: arch-syscall.h endif ifeq ($(subdir),csu) -sysdep_routines += errno-loc +sysdep_routines += errno-loc rseq-sym endif ifeq ($(subdir),assert)
@@ -92,7 +92,8 @@ sysdep_headers += sys/mount.h sys/acct.h \ bits/termios-c_lflag.h bits/termios-tcflow.h \ bits/termios-misc.h \ bits/types/struct_semid_ds.h \ - bits/ipc-perm.h + bits/ipc-perm.h \ + sys/rseq.h bits/rseq.h tests += tst-clone tst-clone2 tst-clone3 tst-fanotify tst-personality \ tst-quota tst-sync_file_range tst-sysconf-iov_max tst-ttyname \
sysdeps/unix/sysv/linux/Versions +1 -0
@@ -178,6 +178,7 @@ libc { getdents64; gettid; tgkill; } GLIBC_2.32 { + __rseq_abi; } GLIBC_PRIVATE { # functions used in other libraries
sysdeps/unix/sysv/linux/aarch64/bits/rseq.h new file +43 -0
43 lines not on this page · raw message
sysdeps/unix/sysv/linux/aarch64/libc.abilist +1 -0
@@ -2149,6 +2149,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/alpha/libc.abilist +1 -0
@@ -2231,6 +2231,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/arm/be/libc.abilist +1 -0
@@ -133,6 +133,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/arm/bits/rseq.h new file +83 -0
83 lines not on this page · raw message
sysdeps/unix/sysv/linux/arm/le/libc.abilist +1 -0
@@ -130,6 +130,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/bits/rseq.h new file +29 -0
@@ -0,0 +1,29 @@ +/* Restartable Sequences architecture header. Stub version. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#ifndef _SYS_RSEQ_H +# error "Never use <bits/rseq.h> directly; include <sys/rseq.h> instead." +#endif + +/* RSEQ_SIG is a signature required before each abort handler code. + +It is a 32-bit value that maps to actual architecture code compiled +into applications and libraries. It needs to be defined for each +architecture. When choosing this value, it needs to be taken into +account that generating invalid instructions may have ill effects on +tools like objdump, and may also have impact on the CPU speculative +execution efficiency in some cases. */
sysdeps/unix/sysv/linux/csky/libc.abilist +1 -0
@@ -2093,6 +2093,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/hppa/libc.abilist +1 -0
@@ -2052,6 +2052,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/i386/libc.abilist +1 -0
@@ -2218,6 +2218,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/ia64/libc.abilist +1 -0
@@ -2084,6 +2084,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/m68k/coldfire/libc.abilist +1 -0
@@ -134,6 +134,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/m68k/m680x0/libc.abilist +1 -0
@@ -2164,6 +2164,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/microblaze/be/libc.abilist +1 -0
@@ -2144,6 +2144,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/microblaze/le/libc.abilist +1 -0
@@ -2141,6 +2141,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/mips/bits/rseq.h new file +62 -0
62 lines not on this page · raw message
sysdeps/unix/sysv/linux/mips/mips32/fpu/libc.abilist +1 -0
@@ -2135,6 +2135,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/mips/mips32/nofpu/libc.abilist +1 -0
@@ -2133,6 +2133,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/mips/mips64/n32/libc.abilist +1 -0
@@ -2141,6 +2141,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/mips/mips64/n64/libc.abilist +1 -0
@@ -2135,6 +2135,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/nios2/libc.abilist +1 -0
@@ -2182,6 +2182,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/powerpc/bits/rseq.h new file +37 -0
@@ -0,0 +1,37 @@ +/* Restartable Sequences Linux powerpc architecture header. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#ifndef _SYS_RSEQ_H +# error "Never use <bits/rseq.h> directly; include <sys/rseq.h> instead." +#endif + +/* RSEQ_SIG is a signature required before each abort handler code. + +It is a 32-bit value that maps to actual architecture code compiled +into applications and libraries. It needs to be defined for each +architecture. When choosing this value, it needs to be taken into +account that generating invalid instructions may have ill effects on +tools like objdump, and may also have impact on the CPU speculative +execution efficiency in some cases. + +RSEQ_SIG uses the following trap instruction: + +powerpc-be: 0f e5 00 0b twui r5,11 +powerpc64-le: 0b 00 e5 0f twui r5,11 +powerpc64-be: 0f e5 00 0b twui r5,11 */ + +#define RSEQ_SIG 0x0fe5000b
sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libc.abilist +1 -0
@@ -2191,6 +2191,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libc.abilist +1 -0
@@ -2224,6 +2224,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/powerpc/powerpc64/be/libc.abilist +1 -0
@@ -2054,6 +2054,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/powerpc/powerpc64/le/libc.abilist +1 -0
@@ -2291,6 +2291,7 @@ GLIBC_2.32 __qecvtieee128_r F GLIBC_2.32 __qfcvtieee128 F GLIBC_2.32 __qfcvtieee128_r F GLIBC_2.32 __qgcvtieee128 F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 __scanfieee128 F GLIBC_2.32 __snprintf_chkieee128 F GLIBC_2.32 __snprintfieee128 F
sysdeps/unix/sysv/linux/riscv/rv64/libc.abilist +1 -0
@@ -2111,6 +2111,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/rseq-internal.h new file +73 -0
73 lines not on this page · raw message
sysdeps/unix/sysv/linux/rseq-sym.c new file +26 -0
@@ -0,0 +1,26 @@ +/* Restartable Sequences exported symbols. Linux Implementation. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#include <sys/syscall.h> +#include <stdint.h> +#include <kernel-features.h> +#include <sys/rseq.h> + +__threadstruct rseq __rseq_abi = + { + .cpu_id =RSEQ_CPU_ID_UNINITIALIZED , + };
sysdeps/unix/sysv/linux/s390/bits/rseq.h new file +37 -0
@@ -0,0 +1,37 @@ +/* Restartable Sequences Linux s390 architecture header. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#ifndef _SYS_RSEQ_H +# error "Never use <bits/rseq.h> directly; include <sys/rseq.h> instead." +#endif + +/* RSEQ_SIG is a signature required before each abort handler code. + +It is a 32-bit value that maps to actual architecture code compiled +into applications and libraries. It needs to be defined for each +architecture. When choosing this value, it needs to be taken into +account that generating invalid instructions may have ill effects on +tools like objdump, and may also have impact on the CPU speculative +execution efficiency in some cases. + +RSEQ_SIG uses the trap4 instruction. As Linux does not make use of the +access-register mode nor the linkage stack this instruction will always +cause a special-operation exception (the trap-enabled bit in the DUCT +is and will stay 0). The instruction pattern is +b2 ff 0f ff trap4 4095(%r0) */ + +#define RSEQ_SIG 0xB2FF0FFF
sysdeps/unix/sysv/linux/s390/s390-32/libc.abilist +1 -0
@@ -2189,6 +2189,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/s390/s390-64/libc.abilist +1 -0
@@ -2090,6 +2090,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/sh/be/libc.abilist +1 -0
@@ -2059,6 +2059,7 @@ GLIBC_2.30 twalk_r F GLIBC_2.31 msgctl F GLIBC_2.31 semctl F GLIBC_2.31 shmctl F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/sh/le/libc.abilist +1 -0
@@ -2056,6 +2056,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/sparc/sparc32/libc.abilist +1 -0
@@ -2180,6 +2180,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/sparc/sparc64/libc.abilist +1 -0
@@ -2107,6 +2107,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/sys/rseq.h new file +228 -0
228 lines not on this page · raw message
sysdeps/unix/sysv/linux/x86/bits/rseq.h new file +30 -0
@@ -0,0 +1,30 @@ +/* Restartable Sequences Linux x86 architecture header. +Copyright (C) 2020 Free Software Foundation, Inc. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Lesser General Public +License as published by the Free Software Foundation; either +version 2.1 of the License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Lesser General Public License for more details. + +You should have received a copy of the GNU Lesser General Public +License along with the GNU C Library; if not, see +<https://www.gnu.org/licenses/>. */ + +#ifndef _SYS_RSEQ_H +# error "Never use <bits/rseq.h> directly; include <sys/rseq.h> instead." +#endif + +/* RSEQ_SIG is a signature required before each abort handler code. + +RSEQ_SIG is used with the following reserved undefined instructions, which +trap in user-space: + +x86-32: 0f b9 3d 53 30 05 53 ud1 0x53053053,%edi +x86-64: 0f b9 3d 53 30 05 53 ud1 0x53053053(%rip),%edi */ + +#define RSEQ_SIG 0x53053053
sysdeps/unix/sysv/linux/x86_64/64/libc.abilist +1 -0
@@ -2065,6 +2065,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
sysdeps/unix/sysv/linux/x86_64/x32/libc.abilist +1 -0
@@ -2162,6 +2162,7 @@ GLIBC_2.30 getdents64 F GLIBC_2.30 gettid F GLIBC_2.30 tgkill F GLIBC_2.30 twalk_r F +GLIBC_2.32 __rseq_abi T 0x20 GLIBC_2.32 pthread_attr_setaffinity_np F GLIBC_2.32 pthread_getaffinity_np F GLIBC_2.32 pthread_getattr_np F
Signature · 1 line
2.17.1
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
* Mathieu Desnoyers:
sysdeps/unix/sysv/linux/sys/rseq.h line 50
Quoted 13 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +13
-0+#ifdef __cplusplus +# if __cplusplus >=201103L +# define __rseq_static_assert (expr, diagnostic) static_assert (expr, diagnostic) +# define __rseq_alignof (type) alignof (type) +# define __rseq_alignas (x) alignas (x) +# define __rseq_tls_storage_class thread_local +# endif +#elif (defined __STDC_VERSION__ ? __STDC_VERSION__ : 0) >= 201112L +# define __rseq_static_assert (expr, diagnostic) _Static_assert (expr, diagnostic) +# define __rseq_alignof (type) _Alignof (type) +# define __rseq_alignas (x) _Alignas (x) +# define __rseq_tls_storage_class _Thread_local +#endif
This does not seem to work. I get this with GCC 9:
In file included from /tmp/cih_test_gsrKbC.cc:8:0:../sysdeps/unix/sysv/linux/sys/rseq.h:42:50: error: attribute ignored [-Werror=attributes]
# define __rseq_alignas(x) alignas (x)
^
../sysdeps/unix/sysv/linux/sys/rseq.h:122:14: note: in expansion of macro ‘__rseq_alignas’
uint32_t __rseq_alignas (32) version;
^In any case, these changes really have to go into the UAPI header first. Only the __thread handling should remain. Otherwise, we'll have a tough situation on our hands changing the UAPI header, without introducing macro definition conflicts. I'd suggest to stick to the aligned attribute for the time being, like the current UAPI headers.
The resut looks okay to me.
I'm still waiting for feedback from other maintainers whether the level of documentation and testing is appropriate.
Thanks,
Florian
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 24 lines · ----- On Jun 3, 2020, at 8:05 AM, Florian Weimer [email protected] wrote:
* Mathieu Desnoyers:
Quoted 11 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +3
-0+#ifdef __cplusplus +# if __cplusplus >=201103L +# define __rseq_static_assert (expr, diagnostic) static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h new file +6
-0+# define __rseq_alignof (type) alignof (type) +# define __rseq_alignas (x) alignas (x) +# define __rseq_tls_storage_class thread_local +# endif +#elif (defined __STDC_VERSION__ ? __STDC_VERSION__ : 0) >= 201112L +# define __rseq_static_assert (expr, diagnostic) _Static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h line 50
Quoted 4 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +4
-0+# define __rseq_alignof (type) _Alignof (type) +# define __rseq_alignas (x) _Alignas (x) +# define __rseq_tls_storage_class _Thread_local +#endif This does not seem to work. I get this with GCC 9:
In file included from
/tmp/cih_test_gsrKbC.cc:8:0:../sysdeps/unix/sysv/linux/sys/rseq.h:42:50: error: attribute ignored [-Werror=attributes]
# define__rseq_alignas(x)alignas (x)
Quoted 5 lines
^
../sysdeps/unix/sysv/linux/sys/rseq.h:122:14: note: in expansion of macro
‘__rseq_alignas’
uint32_t __rseq_alignas (32) version;
^Is that when compiling C or C++ code ? If it's C code, I would expect "_Alignas" to be used, not "alignas".
Which exact version of gcc do you use ?
Quoted 6 lines
In any case, these changes really have to go into the UAPI header first. Only the
__threadhandling should remain. Otherwise, we'll have a tough situation on our hands changing the UAPI header, without introducing macro definition conflicts. I'd suggest to stick to the aligned attribute for the time being, like the current UAPI headers.
OK. Should I do that in a separate patch, or you do it on top of my patchset, or should I re-spin another round of the series ?
Quoted 5 lines
The resut looks okay to me.
I'm still waiting for feedback from other maintainers whether the level of documentation and testing is appropriate.
OK.
Thanks,
Mathieu
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
* Mathieu Desnoyers:
Quoted 32 lines
Quoted 24 lines · ----- On Jun 3, 2020, at 8:05 AM, Florian Weimer [email protected] wrote:
* Mathieu Desnoyers:
Quoted 11 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +3
-0+#ifdef __cplusplus +# if __cplusplus >=201103L +# define __rseq_static_assert (expr, diagnostic) static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h new file +6
-0+# define __rseq_alignof (type) alignof (type) +# define __rseq_alignas (x) alignas (x) +# define __rseq_tls_storage_class thread_local +# endif +#elif (defined __STDC_VERSION__ ? __STDC_VERSION__ : 0) >= 201112L +# define __rseq_static_assert (expr, diagnostic) _Static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h line 50
Quoted 4 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +4
-0+# define __rseq_alignof (type) _Alignof (type) +# define __rseq_alignas (x) _Alignas (x) +# define __rseq_tls_storage_class _Thread_local +#endif This does not seem to work. I get this with GCC 9:
In file included from
/tmp/cih_test_gsrKbC.cc:8:0:../sysdeps/unix/sysv/linux/sys/rseq.h:42:50: error: attribute ignored [-Werror=attributes]
# define__rseq_alignas(x)alignas (x)Quoted 5 lines
^ ../sysdeps/unix/sysv/linux/sys/rseq.h:122:14: note: in expansion of macro ‘__rseq_alignas’ uint32_t __rseq_alignas (32) version; ^
Quoted 4 lines
Is that when compiling C or C++ code ? If it's C code, I would expect "
_Alignas" to be used, not "alignas".Which exact version of gcc do you use ?
C++ code. CXX was set to this compiler at configure time:
gcc version 9.3.1 20200408 (Red Hat 9.3.1-2) (GCC)
Quoted 8 lines
Quoted 5 lines
In any case, these changes really have to go into the UAPI header first. Only the
__threadhandling should remain. Otherwise, we'll have a tough situation on our hands changing the UAPI header, without introducing macro definition conflicts. I'd suggest to stick to the aligned attribute for the time being, like the current UAPI headers.OK. Should I do that in a separate patch, or you do it on top of my patchset, or should I re-spin another round of the series ?
I think the initial commit should mirror the current UAPI header contents.
Keep the macros for the UAPI patch though. 8-) We can pick up these changes once they have been merged into Linux.
Thanks,
Florian
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 34 lines · ----- On Jun 3, 2020, at 8:31 AM, Florian Weimer [email protected] wrote:
* Mathieu Desnoyers:
Quoted 32 lines
Quoted 29 lines · ----- On Jun 3, 2020, at 8:05 AM, Florian Weimer [email protected] wrote:
* Mathieu Desnoyers:
Quoted 11 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +3
-0+#ifdef __cplusplus +# if __cplusplus >=201103L +# define __rseq_static_assert (expr, diagnostic) static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h new file +6
-0+# define __rseq_alignof (type) alignof (type) +# define __rseq_alignas (x) alignas (x) +# define __rseq_tls_storage_class thread_local +# endif +#elif (defined __STDC_VERSION__ ? __STDC_VERSION__ : 0) >= 201112L +# define __rseq_static_assert (expr, diagnostic) _Static_assert (expr,diagnostic)
sysdeps/unix/sysv/linux/sys/rseq.h line 50
Quoted 4 lines
sysdeps/unix/sysv/linux/sys/rseq.h new file +4
-0+# define __rseq_alignof (type) _Alignof (type) +# define __rseq_alignas (x) _Alignas (x) +# define __rseq_tls_storage_class _Thread_local +#endif This does not seem to work. I get this with GCC 9:
In file included from
/tmp/cih_test_gsrKbC.cc:8:0:../sysdeps/unix/sysv/linux/sys/rseq.h:42:50: error: attribute ignored [-Werror=attributes]
# define__rseq_alignas(x)alignas (x)^ ../sysdeps/unix/sysv/linux/sys/rseq.h:122:14: note: in expansion of macro ‘__rseq_alignas’ uint32_t __rseq_alignas (32) version; ^
Quoted 8 lines
Quoted 4 lines
Is that when compiling C or C++ code ? If it's C code, I would expect "
_Alignas" to be used, not "alignas".Which exact version of gcc do you use ?
C++ code. CXX was set to this compiler at configure time:
gcc version 9.3.1 20200408 (Red Hat 9.3.1-2) (GCC)
I think I found the culprit: it should be:
__rseq_alignas (32) uint32_t version;
rather than the other way around.
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
In any case, these changes really have to go into the UAPI header first. Only the
__threadhandling should remain. Otherwise, we'll have a tough situation on our hands changing the UAPI header, without introducing macro definition conflicts. I'd suggest to stick to the aligned attribute for the time being, like the current UAPI headers.
OK, so I just remove the __rseq_alignas for now and use "aligned()" instead like the UAPI header. I plan to keep the other macros for now.
Quoted 9 lines
Quoted 3 lines
OK. Should I do that in a separate patch, or you do it on top of my patchset, or should I re-spin another round of the series ?
I think the initial commit should mirror the current UAPI header contents.
Keep the macros for the UAPI patch though. 8-) We can pick up these changes once they have been merged into Linux.
OK,
Thanks!
Mathieu
Quoted 3 lines
Thanks,
Florian
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
In reply to Florian WeimerQuoted 2 lines · Florian Weimer via Libc-alpha
I'm still waiting for feedback from other maintainers whether the level of documentation and testing is appropriate.
Looking at the documentation in the manual, it doesn't look like it has enough information for someone to use this functionality, or to know when they might want to use it, and nor does it point to external documentation of it. It would seem appropriate at least to include a link to wherever the external documentation is of what this functionality is good for, how to create and use a "Restartable Sequence critical section", and how to "perform rseq registration to the kernel".
Signature · 2 lines
Joseph S. Myers [email protected]
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 5 lines · ----- On Jun 3, 2020, at 1:22 PM, Joseph Myers [email protected] wrote:
Quoted 2 lines · Florian Weimer via Libc-alpha
I'm still waiting for feedback from other maintainers whether the level of documentation and testing is appropriate.
Quoted 7 lines
Looking at the documentation in the manual, it doesn't look like it has enough information for someone to use this functionality, or to know when they might want to use it, and nor does it point to external documentation of it. It would seem appropriate at least to include a link to wherever the external documentation is of what this functionality is good for, how to create and use a "Restartable Sequence critical section", and how to "perform rseq registration to the kernel".
That external piece of documentation would be part of the Linux man-pages project, maintained by Michael Kerrisk. I have submitted a few revisions of the rseq(2) man page, but have been waiting for Michael to reply for more than a year now:
https://lore.kernel.org/r/[email protected]
I'm thinking about hosting a rseq(2) man-page into my librseq project, would that make sense ?
librseq is currently a development branch in my own repository for now. It still needs to be tweaked to adapt to the various changes that went into the glibc rseq enablement patchset, and then I plan to move it to a more "formal" home:
https://github.com/compudj/librseq
Thanks,
Mathieu
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 9 lines · Mathieu Desnoyers via Libc-alpha
That external piece of documentation would be part of the Linux man-pages project, maintained by Michael Kerrisk. I have submitted a few revisions of the rseq(2) man page, but have been waiting for Michael to reply for more than a year now:
https://lore.kernel.org/r/[email protected]
I'm thinking about hosting a rseq(2) man-page into my librseq project, would that make sense ?
I'm not particularly concerned with exactly where it goes, as long as it's somewhere stable we can link to.
Signature · 2 lines
Joseph S. Myers [email protected]
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 7 lines · ----- On Jun 4, 2020, at 1:46 PM, Joseph Myers [email protected] wrote:
Quoted 5 lines · Mathieu Desnoyers via Libc-alpha
That external piece of documentation would be part of the Linux man-pages project, maintained by Michael Kerrisk. I have submitted a few revisions of the rseq(2) man page, but have been waiting for Michael to reply for more than a year now:
Quoted 7 lines
Quoted 4 lines
https://lore.kernel.org/r/[email protected]
I'm thinking about hosting a rseq(2) man-page into my librseq project, would that make sense ?
I'm not particularly concerned with exactly where it goes, as long as it's somewhere stable we can link to.
I managed to get a repository up and running for librseq, and have integrated the rseq.2 man page with comments from Michael Kerrisk here:
https://git.kernel.org/pub/scm/libs/librseq/librseq.git/tree/doc/man/rseq.2
Is that a suitable URL ? Can we simply point to it from glibc's manual ?
Thanks,
Mathieu
Quoted 4 lines
--
Joseph S. Myers
[email protected]
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 6 lines · Mathieu Desnoyers
I managed to get a repository up and running for librseq, and have integrated the rseq.2 man page with comments from Michael Kerrisk here:
https://git.kernel.org/pub/scm/libs/librseq/librseq.git/tree/doc/man/rseq.2
Is that a suitable URL ? Can we simply point to it from glibc's manual ?
Yes, that seems something reasonable to link to.
Signature · 2 lines
Joseph S. Myers [email protected]
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 9 lines · Joseph Myers
Quoted 6 lines · Mathieu Desnoyers
I managed to get a repository up and running for librseq, and have integrated the rseq.2 man page with comments from Michael Kerrisk here:
https://git.kernel.org/pub/scm/libs/librseq/librseq.git/tree/doc/man/rseq.2
Is that a suitable URL ? Can we simply point to it from glibc's manual ?
Yes, that seems something reasonable to link to.
is there work to make the usage of rseq critical sections portable? (e.g. transactional memory critical section has syntax in gcc, but that doesn't require straight line code with begin/end/abort labels in a particular layout.)
the macros and inline asm in rseq-*.h are not too nice, but if they can completely hide the non-portable bits then i guess that works.
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
Quoted 17 lines · ----- On Jun 18, 2020, at 8:22 AM, Szabolcs Nagy [email protected] wrote:
Quoted 9 lines · Joseph Myers
Quoted 6 lines · Mathieu Desnoyers
I managed to get a repository up and running for librseq, and have integrated the rseq.2 man page with comments from Michael Kerrisk here:
https://git.kernel.org/pub/scm/libs/librseq/librseq.git/tree/doc/man/rseq.2
Is that a suitable URL ? Can we simply point to it from glibc's manual ?
Yes, that seems something reasonable to link to.
is there work to make the usage of rseq critical sections portable? (e.g. transactional memory critical section has syntax in gcc, but that doesn't require straight line code with begin/end/abort labels in a particular layout.)
Quoted 3 lines
the macros and inline asm in rseq-*.h are not too nice, but if they can completely hide the non-portable bits then i guess that works.
My goal with librseq is indeed to provide static inlines which hide the architecture-specific ugliness of rseq critical section assembly code behind an API which can be used from all supported architectures for most of the known use-cases, so only very specific use-case would have to craft their own assembly.
Thanks,
Mathieu
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Re: [PATCH glibc 1/3] glibc: Perform rseq registration at C startup and thread creation (v20)
In reply to Joseph MyersQuoted 10 lines · ----- On Jun 11, 2020, at 4:26 PM, Joseph Myers [email protected] wrote:
Quoted 6 lines · Mathieu Desnoyers
I managed to get a repository up and running for librseq, and have integrated the rseq.2 man page with comments from Michael Kerrisk here:
https://git.kernel.org/pub/scm/libs/librseq/librseq.git/tree/doc/man/rseq.2
Is that a suitable URL ? Can we simply point to it from glibc's manual ?
Yes, that seems something reasonable to link to.
OK, I'll point to that then. I've been waiting for Michael Kerrisk to act on rseq(2) for more than a year now, and this is getting in the way of upstreaming rseq support into glibc 2.32.
Thanks,
Mathieu
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
PATCH 2/3 Linux: Use rseq in sched_getcpu if available (v9)
In reply to a message not in the archiveWhen available, use the cpu_id field from __rseq_abi on Linux to implement sched_getcpu(). Fall-back on the vgetcpu vDSO if unavailable.
Benchmarks:
x86-64: Intel E5-2630 [email protected], 16-core, hyperthreading
glibc sched_getcpu(): 13.7 ns (baseline) glibc sched_getcpu() using rseq: 2.5 ns (speedup: 5.5x)
inline load cpuid from __rseq_abi TLS: 0.8 ns (speedup: 17.1x)
---
Changes since v1:
- rseq is only used if both __NR_rseq and RSEQ_SIG are defined.
Changes since v2:
- remove duplicated __rseq_abi extern declaration.
Changes since v3:
- update ChangeLog.
Changes since v4:
- Use atomic_load_relaxed to load the __rseq_abi.cpu_id field, a consequence of the fact that __rseq_abi is not volatile anymore.
- Include atomic.h which provides atomic_load_relaxed.
Changes since v5:
- Use __ASSUME_RSEQ to detect rseq availability.
Changes since v6:
- Remove use of __ASSUME_RSEQ.
Changes since v7:
- Fix incorrect merge with commit d0def09ff6 ("linux: Fix vDSO macros build with time64 interfaces")
Changes since v8:
- Update patch title.
- Add /* RSEQ_SIG */ for #else and #endif.
--- sysdeps/unix/sysv/linux/sched_getcpu.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-)
sysdeps/unix/sysv/linux/sched_getcpu.c +20 -2
@@ -18,10 +18,12 @@#include <errno.h> #include <sched.h> #include <sysdep.h> +#include <atomic.h> #include <sysdep-vdso.h> +#include <sys/rseq.h> -int -sched_getcpu (void ) +static int +vsyscall_sched_getcpu (void ) {unsigned int cpu;int r =-1 ;
@@ -32,3 +34,19 @@ sched_getcpu (void)#endif return r ==-1 ? r : cpu; } + +#ifdef RSEQ_SIG +int +sched_getcpu (void ) +{ +int cpu_id =atomic_load_relaxed (&__rseq_abi.cpu_id); + +return cpu_id >=0 ? cpu_id :vsyscall_sched_getcpu (); +} +#else /* RSEQ_SIG */ +int +sched_getcpu (void ) +{ +return vsyscall_sched_getcpu (); +} +#endif /* RSEQ_SIG */
Signature · 1 line
2.17.1