From patchwork Thu Nov 23 21:40:21 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Michael Jeanson X-Patchwork-Id: 80668 Return-Path: X-Original-To: patchwork@sourceware.org Delivered-To: patchwork@sourceware.org Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 8EB2C3858C5F for ; Thu, 23 Nov 2023 21:40:54 +0000 (GMT) X-Original-To: libc-alpha@sourceware.org Delivered-To: libc-alpha@sourceware.org Received: from smtpout.efficios.com (unknown [IPv6:2607:5300:203:b2ee::31e5]) by sourceware.org (Postfix) with ESMTPS id 15FB43858D1E for ; Thu, 23 Nov 2023 21:40:39 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 15FB43858D1E Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=efficios.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=efficios.com ARC-Filter: OpenARC Filter v1.0.0 sourceware.org 15FB43858D1E Authentication-Results: server2.sourceware.org; arc=none smtp.remote-ip=2607:5300:203:b2ee::31e5 ARC-Seal: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1700775643; cv=none; b=rGzme248UTPv61RSpnTDsi5oD8vR2lEDKZx14vFoWiymMHxmhGzdLbyZpRLPqQLNHSlmHieUQX6Rp5OFUCPakw51LSQxEBfMZbK+GlP6YkLQ9smYOXHogad1DORghgLiui3W+9aHWFPxTRGtqc23RPvXNNFEHKiE/57tXQCpLtw= ARC-Message-Signature: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1700775643; c=relaxed/simple; bh=GQ+QWvm9pWthIdJlRGuW5E6RACXK7DeQT/fMKG9ZtkU=; h=DKIM-Signature:From:To:Subject:Date:Message-Id:MIME-Version; b=GqA12VhKXiB695dIPPQgxtGqkL9+NZ5SDLpFNYMIU4jVmSkBIsMzXv9eZFc2y6lWeWO2dfXD0a1NUfCkDtCLXAgYkcyTDqgm1TKFLpwzT6tCpvOqzmqj+fX0H4vIawAUi9HJfk7GJ8VpDJHyn8RzaVYpQkXVTLqFibxfsCrkeu8= ARC-Authentication-Results: i=1; server2.sourceware.org DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=efficios.com; s=smtpout1; t=1700775638; bh=GQ+QWvm9pWthIdJlRGuW5E6RACXK7DeQT/fMKG9ZtkU=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=xeDoC3abeV7FT/jkyGvyJ4zn6V2aexiWe2lF/Ei7r6zJprwqpNSWKCD9f5swzyGAf CH/1F94+Wi5dfm6kuBM5XvoREsFjWGz7BZ8kPeMe0QIGNlp56iTYPj7X0i1LZZp00B dzVvYQRFzsLN2DNjDSAJtJXILzTQypPENfDKdzoX1zq094QPHplXYMVgXGXnJMz9rm AFZ87+BHMHkI6mjCD6FUu3MJCemeoaGkwSrxIaORKWO0gO11cYEYB0dhatZLMPmXow hRfioB9GtIKM/wtkhY/avezvu5jiIW1b35HzV7vYUHrlnEx8k1GiS8bvZdgbtQz9cb 3U7kLG9WddQ6A== Received: from laptop-mjeanson.internal.efficios.com (192-222-143-198.qc.cable.ebox.net [192.222.143.198]) by smtpout.efficios.com (Postfix) with ESMTPSA id 4Sbs3p1lj1z1d2Q; Thu, 23 Nov 2023 16:40:38 -0500 (EST) From: Michael Jeanson To: libc-alpha@sourceware.org Cc: Michael Jeanson , Mathieu Desnoyers , Florian Weimer , Carlos O'Donell Subject: [RFC PATCH v2] Add rseq extensible ABI support Date: Thu, 23 Nov 2023 16:40:21 -0500 Message-Id: <20231123214021.286351-1-mjeanson@efficios.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231122214020.247888-1-mjeanson@efficios.com> References: <20231122214020.247888-1-mjeanson@efficios.com> MIME-Version: 1.0 X-Spam-Status: No, score=-12.2 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, GIT_PATCH_0, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: libc-alpha@sourceware.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Libc-alpha mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libc-alpha-bounces+patchwork=sourceware.org@sourceware.org Introduced in Linux v6.3 the rseq extensible ABI [1] will allow adding rseq features past the initial 32 bytes of the original ABI. While the rseq features in the latest kernel still fit within the original ABI size, there are currently only 4 bytes left. It would thus be a good time to add support for the extensible ABI so that when new features are added, they are immediately available to GNU libc users. This is a first draft of a possible implementation to gather some feedback, it currently only supports dynamically linked executables. We use the ELF auxiliary vectors to query the kernel for the size and alignment of the rseq area, if this fails we default to the original fixed size and alignment of '32' which the kernel will accept as a compatibility mode with the original ABI. This makes the size of the rseq area variable and thus requires to relocate it out of 'struct pthread'. We chose to move it after (in block allocation order) the last TLS block since it required a fairly small modification to the TLS block allocator and did not interfere with the main executable TLS block which must always be first. [1] https://lore.kernel.org/all/20221122203932.231377-4-mathieu.desnoyers@efficios.com/ Signed-off-by: Michael Jeanson Co-Authored-By: Mathieu Desnoyers Signed-off-By: Mathieu Desnoyers Cc: Florian Weimer Cc: Carlos O'Donell --- Changes since RFC v1: - Insert the rseq area after the last TLS block - Add proper support for TLS_TCB_AT_TP variant --- csu/libc-tls.c | 6 +++ elf/dl-tls.c | 56 +++++++++++++++++++++++++ elf/rtld_static_init.c | 6 +++ nptl/descr.h | 20 ++------- nptl/pthread_create.c | 2 +- sysdeps/generic/ldsodefs.h | 6 +++ sysdeps/i386/nptl/tcb-access.h | 54 ++++++++++++++++++++++++ sysdeps/nptl/dl-tls_init_tp.c | 4 +- sysdeps/nptl/tcb-access.h | 5 +++ sysdeps/unix/sysv/linux/rseq-internal.h | 26 ++++++++++-- sysdeps/unix/sysv/linux/sched_getcpu.c | 3 +- sysdeps/unix/sysv/linux/tst-rseq.c | 10 +++-- sysdeps/unix/sysv/linux/tst-rseq.h | 3 +- sysdeps/x86_64/nptl/tcb-access.h | 55 +++++++++++++++++++++++- 14 files changed, 225 insertions(+), 31 deletions(-) diff --git a/csu/libc-tls.c b/csu/libc-tls.c index cdf6442c02..8458f47577 100644 --- a/csu/libc-tls.c +++ b/csu/libc-tls.c @@ -62,6 +62,12 @@ size_t _dl_tls_static_surplus; dynamic TLS access (e.g. with TLSDESC). */ size_t _dl_tls_static_optional; +/* Offset of the rseq area from the thread pointer. */ +ptrdiff_t _dl_tls_rseq_offset; + +/* Size of the rseq area in the static TLS block. */ +size_t _dl_tls_rseq_size; + /* Generation counter for the dtv. */ size_t _dl_tls_generation; diff --git a/elf/dl-tls.c b/elf/dl-tls.c index c192b5a13a..e9390f87fd 100644 --- a/elf/dl-tls.c +++ b/elf/dl-tls.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include @@ -75,6 +76,12 @@ /* Default for dl_tls_static_optional. */ #define OPTIONAL_TLS 512 +/* Minimum size of the rseq area. */ +#define TLS_DL_RSEQ_MIN_SIZE 32 + +/* Minimum size of the rseq area alignment. */ +#define TLS_DL_RSEQ_MIN_ALIGN 32 + /* Compute the static TLS surplus based on the namespace count and the TLS space that can be used for optimizations. */ static inline int @@ -297,6 +304,29 @@ _dl_determine_tlsoffset (void) slotinfo[cnt].map->l_tls_offset = off; } + /* Insert the rseq area block after the last TLS block. */ + + /* Get the rseq auxiliary vectors, 0 is returned when not implemented + and we then default to the rseq ABI minimums. */ + size_t rseq_size = MAX (getauxval (AT_RSEQ_FEATURE_SIZE), TLS_DL_RSEQ_MIN_SIZE); + size_t rseq_align = MAX (getauxval (AT_RSEQ_ALIGN), TLS_DL_RSEQ_MIN_ALIGN); + + /* Make sure the rseq area size is a multiple of the requested + aligment. */ + rseq_size = roundup (rseq_size, rseq_align); + + /* Add the rseq area block to the global offset. */ + offset = roundup (offset, rseq_align) + rseq_size; + + /* Increase the max_align if necessary. */ + max_align = MAX (max_align, rseq_align); + + /* Record the rseq_area block size and offset. The offset is negative + with TLS_TCB_AT_TP because the TLS blocks are located before the + thread pointer. */ + GLRO (dl_tls_rseq_offset) = -offset; + GLRO (dl_tls_rseq_size) = rseq_size; + GL(dl_tls_static_used) = offset; GLRO (dl_tls_static_size) = (roundup (offset + GLRO(dl_tls_static_surplus), max_align) @@ -342,6 +372,32 @@ _dl_determine_tlsoffset (void) offset = off + slotinfo[cnt].map->l_tls_blocksize - firstbyte; } + /* Insert the rseq area block after the last TLS block. */ + + /* Get the rseq auxiliary vectors, 0 is returned when not implemented + and we then default to the rseq ABI minimums. */ + size_t rseq_size = MAX (getauxval (AT_RSEQ_FEATURE_SIZE), TLS_DL_RSEQ_MIN_SIZE); + size_t rseq_align = MAX (getauxval (AT_RSEQ_ALIGN), TLS_DL_RSEQ_MIN_ALIGN); + + /* Make sure the rseq area size is a multiple of the requested + aligment. */ + rseq_size = roundup (rseq_size, rseq_align); + + /* Align the global offset to the beginning of the rseq area. */ + offset = roundup (offset, rseq_align); + + /* Record the rseq_area block size and offset. The offset is positive + with TLS_DTV_AT_TP because the TLS blocks are located after the + thread pointer. */ + GLRO (dl_tls_rseq_size) = rseq_size; + GLRO (dl_tls_rseq_offset) = offset; + + /* Add the rseq area block to the global offset. */ + offset += rseq_size; + + /* Increase the max_align if necessary. */ + max_align = MAX (max_align, rseq_align); + GL(dl_tls_static_used) = offset; GLRO (dl_tls_static_size) = roundup (offset + GLRO(dl_tls_static_surplus), TCB_ALIGNMENT); diff --git a/elf/rtld_static_init.c b/elf/rtld_static_init.c index aec8cc056b..84548023ab 100644 --- a/elf/rtld_static_init.c +++ b/elf/rtld_static_init.c @@ -78,6 +78,12 @@ __rtld_static_init (struct link_map *map) extern __typeof (dl->_dl_tls_static_size) _dl_tls_static_size attribute_hidden; dl->_dl_tls_static_size = _dl_tls_static_size; + extern __typeof (dl->_dl_tls_rseq_size) _dl_tls_rseq_size + attribute_hidden; + dl->_dl_tls_rseq_size = _dl_tls_rseq_size; + extern __typeof (dl->_dl_tls_rseq_offset) _dl_tls_rseq_offset + attribute_hidden; + dl->_dl_tls_rseq_offset = _dl_tls_rseq_offset; dl->_dl_find_object = _dl_find_object; __rtld_static_init_arch (map, dl); diff --git a/nptl/descr.h b/nptl/descr.h index 0171576c23..fadba1f619 100644 --- a/nptl/descr.h +++ b/nptl/descr.h @@ -404,25 +404,11 @@ struct pthread /* Used on strsignal. */ struct tls_internal_t tls_state; - /* rseq area registered with the kernel. Use a custom definition - here to isolate from kernel struct rseq changes. The - implementation of sched_getcpu needs acccess to the cpu_id field; - the other fields are unused and not included here. */ - union - { - struct - { - uint32_t cpu_id_start; - uint32_t cpu_id; - }; - char pad[32]; /* Original rseq area size. */ - } rseq_area __attribute__ ((aligned (32))); - /* Amount of end padding, if any, in this structure. - This definition relies on rseq_area being last. */ + This definition relies on tls_state being last. */ #define PTHREAD_STRUCT_END_PADDING \ - (sizeof (struct pthread) - offsetof (struct pthread, rseq_area) \ - + sizeof ((struct pthread) {}.rseq_area)) + (sizeof (struct pthread) - offsetof (struct pthread, tls_state) \ + + sizeof ((struct pthread) {}.tls_state)) } __attribute ((aligned (TCB_ALIGNMENT))); static inline bool diff --git a/nptl/pthread_create.c b/nptl/pthread_create.c index 63cb684f04..4ee0f24741 100644 --- a/nptl/pthread_create.c +++ b/nptl/pthread_create.c @@ -691,7 +691,7 @@ __pthread_create_2_1 (pthread_t *newthread, const pthread_attr_t *attr, /* Inherit rseq registration state. Without seccomp filters, rseq registration will either always fail or always succeed. */ - if ((int) THREAD_GETMEM_VOLATILE (self, rseq_area.cpu_id) >= 0) + if ((int) RSEQ_GETMEM_VOLATILE (rseq_get_area(), cpu_id) >= 0) pd->flags |= ATTR_FLAG_DO_RSEQ; /* Initialize the field for the ID of the thread which is waiting diff --git a/sysdeps/generic/ldsodefs.h b/sysdeps/generic/ldsodefs.h index 9b50ddd09f..f8a3628bc0 100644 --- a/sysdeps/generic/ldsodefs.h +++ b/sysdeps/generic/ldsodefs.h @@ -610,6 +610,12 @@ struct rtld_global_ro See comments in elf/dl-tls.c where it is initialized. */ EXTERN size_t _dl_tls_static_surplus; + /* Offset of the rseq area from the thread pointer. */ + EXTERN ptrdiff_t _dl_tls_rseq_offset; + + /* Size of the rseq area in the static TLS block. */ + EXTERN size_t _dl_tls_rseq_size; + /* Name of the shared object to be profiled (if any). */ EXTERN const char *_dl_profile; /* Filename of the output file. */ diff --git a/sysdeps/i386/nptl/tcb-access.h b/sysdeps/i386/nptl/tcb-access.h index 28f0a5625f..f1bbccab6e 100644 --- a/sysdeps/i386/nptl/tcb-access.h +++ b/sysdeps/i386/nptl/tcb-access.h @@ -123,3 +123,57 @@ "i" (offsetof (struct pthread, member)), \ "r" (idx)); \ }}) + + +/* Read member of the RSEQ area directly. */ +#define RSEQ_GETMEM_VOLATILE(descr, member) \ + ({ __typeof (descr->member) __value; \ + _Static_assert (sizeof (__value) == 1 \ + || sizeof (__value) == 4 \ + || sizeof (__value) == 8, \ + "size of per-thread data"); \ + if (sizeof (__value) == 1) \ + asm volatile ("movb %%gs:%P2(%3),%b0" \ + : "=q" (__value) \ + : "0" (0), "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else if (sizeof (__value) == 4) \ + asm volatile ("movl %%gs:%P1(%2),%0" \ + : "=r" (__value) \ + : "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else /* 8 */ \ + { \ + asm volatile ("movl %%gs:%P1(%2),%%eax\n\t" \ + "movl %%gs:4+%P1(%2),%%edx" \ + : "=&A" (__value) \ + : "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + } \ + __value; }) + +/* Set member of the RSEQ area directly. */ +#define RSEQ_SETMEM(descr, member, value) \ + ({ \ + _Static_assert (sizeof (descr->member) == 1 \ + || sizeof (descr->member) == 4 \ + || sizeof (descr->member) == 8, \ + "size of per-thread data"); \ + if (sizeof (descr->member) == 1) \ + asm volatile ("movb %b0,%%gs:%P1(%2)" : \ + : "iq" (value), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else if (sizeof (descr->member) == 4) \ + asm volatile ("movl %0,%%gs:%P1(%2)" : \ + : "ir" (value), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else /* 8 */ \ + { \ + asm volatile ("movl %%eax,%%gs:%P1(%2)\n\t" \ + "movl %%edx,%%gs:4+%P1(%2)" : \ + : "A" ((uint64_t) cast_to_integer (value)), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + }}) diff --git a/sysdeps/nptl/dl-tls_init_tp.c b/sysdeps/nptl/dl-tls_init_tp.c index 2ed98c5a31..406f6ba716 100644 --- a/sysdeps/nptl/dl-tls_init_tp.c +++ b/sysdeps/nptl/dl-tls_init_tp.c @@ -108,7 +108,7 @@ __tls_init_tp (void) /* We need a writable view of the variables. They are in .data.relro and are not yet write-protected. */ extern unsigned int size __asm__ ("__rseq_size"); - size = sizeof (pd->rseq_area); + size = GLRO (dl_tls_rseq_size); } #ifdef RSEQ_SIG @@ -118,7 +118,7 @@ __tls_init_tp (void) if the rseq registration may have happened because RSEQ_SIG is defined. */ extern ptrdiff_t offset __asm__ ("__rseq_offset"); - offset = (char *) &pd->rseq_area - (char *) __thread_pointer (); + offset = GLRO (dl_tls_rseq_offset); #endif } diff --git a/sysdeps/nptl/tcb-access.h b/sysdeps/nptl/tcb-access.h index d6dfd41391..7092428d38 100644 --- a/sysdeps/nptl/tcb-access.h +++ b/sysdeps/nptl/tcb-access.h @@ -30,3 +30,8 @@ descr->member = (value) #define THREAD_SETMEM_NC(descr, member, idx, value) \ descr->member[idx] = (value) + +#define RSEQ_GETMEM_VOLATILE(descr, member) \ + THREAD_GETMEM_VOLATILE(descr, member) +#define RSEQ_SETMEM(descr, member, value) \ + THREAD_SETMEM(descr, member, value) diff --git a/sysdeps/unix/sysv/linux/rseq-internal.h b/sysdeps/unix/sysv/linux/rseq-internal.h index 294880c04e..a944c71421 100644 --- a/sysdeps/unix/sysv/linux/rseq-internal.h +++ b/sysdeps/unix/sysv/linux/rseq-internal.h @@ -24,6 +24,24 @@ #include #include #include +#include +#include + +/* rseq area registered with the kernel. Use a custom definition + here to isolate from kernel struct rseq changes. The + implementation of sched_getcpu needs acccess to the cpu_id field; + the other fields are unused and not included here. */ +struct rseq_area +{ + uint32_t cpu_id_start; + uint32_t cpu_id; +}; + +static inline struct rseq_area * +rseq_get_area(void) +{ + return (struct rseq_area *) ((char *) __thread_pointer() + GLRO (dl_tls_rseq_offset)); +} #ifdef RSEQ_SIG static inline bool @@ -31,20 +49,20 @@ rseq_register_current_thread (struct pthread *self, bool do_rseq) { if (do_rseq) { - int ret = INTERNAL_SYSCALL_CALL (rseq, &self->rseq_area, - sizeof (self->rseq_area), + int ret = INTERNAL_SYSCALL_CALL (rseq, rseq_get_area(), + GLRO (dl_tls_rseq_size), 0, RSEQ_SIG); if (!INTERNAL_SYSCALL_ERROR_P (ret)) return true; } - THREAD_SETMEM (self, rseq_area.cpu_id, RSEQ_CPU_ID_REGISTRATION_FAILED); + RSEQ_SETMEM (rseq_get_area(), cpu_id, RSEQ_CPU_ID_REGISTRATION_FAILED); return false; } #else /* RSEQ_SIG */ static inline bool rseq_register_current_thread (struct pthread *self, bool do_rseq) { - THREAD_SETMEM (self, rseq_area.cpu_id, RSEQ_CPU_ID_REGISTRATION_FAILED); + RSEQ_SETMEM (rseq_get_area(), cpu_id, RSEQ_CPU_ID_REGISTRATION_FAILED); return false; } #endif /* RSEQ_SIG */ diff --git a/sysdeps/unix/sysv/linux/sched_getcpu.c b/sysdeps/unix/sysv/linux/sched_getcpu.c index 4457d714bc..6109c68625 100644 --- a/sysdeps/unix/sysv/linux/sched_getcpu.c +++ b/sysdeps/unix/sysv/linux/sched_getcpu.c @@ -19,6 +19,7 @@ #include #include #include +#include static int vsyscall_sched_getcpu (void) @@ -37,7 +38,7 @@ vsyscall_sched_getcpu (void) int sched_getcpu (void) { - int cpu_id = THREAD_GETMEM_VOLATILE (THREAD_SELF, rseq_area.cpu_id); + int cpu_id = RSEQ_GETMEM_VOLATILE (rseq_get_area(), cpu_id); return __glibc_likely (cpu_id >= 0) ? cpu_id : vsyscall_sched_getcpu (); } #else /* RSEQ_SIG */ diff --git a/sysdeps/unix/sysv/linux/tst-rseq.c b/sysdeps/unix/sysv/linux/tst-rseq.c index 16983503b1..a6af0e783f 100644 --- a/sysdeps/unix/sysv/linux/tst-rseq.c +++ b/sysdeps/unix/sysv/linux/tst-rseq.c @@ -31,18 +31,20 @@ # include # include # include +# include # include "tst-rseq.h" static void do_rseq_main_test (void) { - struct pthread *pd = THREAD_SELF; + size_t rseq_size = MAX (getauxval (AT_RSEQ_FEATURE_SIZE), 32); TEST_VERIFY_EXIT (rseq_thread_registered ()); TEST_COMPARE (__rseq_flags, 0); - TEST_VERIFY ((char *) __thread_pointer () + __rseq_offset - == (char *) &pd->rseq_area); - TEST_COMPARE (__rseq_size, sizeof (pd->rseq_area)); + //FIXME: unsure how to test this + //TEST_VERIFY ((char *) __thread_pointer () + __rseq_offset + // == (char *) &pd->rseq_area); + TEST_COMPARE (__rseq_size, rseq_size); } static void diff --git a/sysdeps/unix/sysv/linux/tst-rseq.h b/sysdeps/unix/sysv/linux/tst-rseq.h index 95d12048df..82ce90b62c 100644 --- a/sysdeps/unix/sysv/linux/tst-rseq.h +++ b/sysdeps/unix/sysv/linux/tst-rseq.h @@ -23,11 +23,12 @@ #include #include #include +#include static inline bool rseq_thread_registered (void) { - return THREAD_GETMEM_VOLATILE (THREAD_SELF, rseq_area.cpu_id) >= 0; + return RSEQ_GETMEM_VOLATILE (rseq_get_area(), cpu_id) >= 0; } static inline int diff --git a/sysdeps/x86_64/nptl/tcb-access.h b/sysdeps/x86_64/nptl/tcb-access.h index 110b1be44d..d45413b1e3 100644 --- a/sysdeps/x86_64/nptl/tcb-access.h +++ b/sysdeps/x86_64/nptl/tcb-access.h @@ -67,7 +67,6 @@ } \ __value; }) - /* Loading addresses of objects on x86-64 needs to be treated special when generating PIC code. */ #ifdef __pic__ @@ -130,3 +129,57 @@ "i" (offsetof (struct pthread, member[0])), \ "r" (idx)); \ }}) + +/* Read member of the RSEQ area directly. */ +# define RSEQ_GETMEM_VOLATILE(descr, member) \ + ({ __typeof (descr->member) __value; \ + _Static_assert (sizeof (__value) == 1 \ + || sizeof (__value) == 4 \ + || sizeof (__value) == 8, \ + "size of per-thread data"); \ + if (sizeof (__value) == 1) \ + asm volatile ("movb %%fs:%P2(%q3),%b0" \ + : "=q" (__value) \ + : "0" (0), "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else if (sizeof (__value) == 4) \ + asm volatile ("movl %%fs:%P1(%q2),%0" \ + : "=r" (__value) \ + : "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else /* 8 */ \ + { \ + asm volatile ("movq %%fs:%P1(%q2),%q0" \ + : "=r" (__value) \ + : "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + } \ + __value; }) + +/* Set member of the RSEQ area directly. */ +# define RSEQ_SETMEM(descr, member, value) \ + ({ \ + _Static_assert (sizeof (descr->member) == 1 \ + || sizeof (descr->member) == 4 \ + || sizeof (descr->member) == 8, \ + "size of per-thread data"); \ + if (sizeof (descr->member) == 1) \ + asm volatile ("movb %b0,%%fs:%P1(%q2)" : \ + : "iq" (value), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else if (sizeof (descr->member) == 4) \ + asm volatile ("movl %0,%%fs:%P1(%q2)" : \ + : IMM_MODE (value), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + else /* 8 */ \ + { \ + /* Since movq takes a signed 32-bit immediate or a register source \ + operand, use "er" constraint for 32-bit signed integer constant \ + or register. */ \ + asm volatile ("movq %q0,%%fs:%P1(%q2)" : \ + : "er" ((uint64_t) cast_to_integer (value)), \ + "i" (offsetof (struct rseq_area, member)), \ + "r" (__rseq_offset)); \ + }})