From patchwork Mon Oct 14 20:56:42 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Adhemerval Zanella X-Patchwork-Id: 1176638 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=sourceware.org (client-ip=209.132.180.131; helo=sourceware.org; envelope-from=libc-alpha-return-105932-incoming=patchwork.ozlabs.org@sourceware.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=linaro.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; secure) header.d=sourceware.org header.i=@sourceware.org header.b="qpwRgDm3"; dkim=pass (2048-bit key; unprotected) header.d=linaro.org header.i=@linaro.org header.b="ob+GoW4Q"; dkim-atps=neutral Received: from sourceware.org (server1.sourceware.org [209.132.180.131]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 46sW9W5XKfz9sP6 for ; Tue, 15 Oct 2019 07:58:07 +1100 (AEDT) DomainKey-Signature: a=rsa-sha1; c=nofws; d=sourceware.org; h=list-id :list-unsubscribe:list-subscribe:list-archive:list-post :list-help:sender:from:to:subject:date:message-id:in-reply-to :references; q=dns; s=default; b=qR3tH3a8kgziIXbc9+m4b94X+wvsOxg krffK3ViVPzQKCbPelswD9okdy+1LHMubrvkNgABYz8bHCiIUexiLq4G8VIg8Lxi TaS8ZJAD5h4gdHPjpJi7h+Su/AvH6vDhQew0cMfVvAmyrLSpQ0M267zNM/wtS+Eb MYxRIX+ac9hk= DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=sourceware.org; h=list-id :list-unsubscribe:list-subscribe:list-archive:list-post :list-help:sender:from:to:subject:date:message-id:in-reply-to :references; s=default; bh=Wm/2VyhDJnZDUOZ6lHjYPog/aOI=; b=qpwRg Dm3JeRdZdYzTKOthIWB0c0j7FEmSkGeOszuM0b6Vw4mwHvWnD+bmNXxQ1INyIXEe wODSjAwwFOPIZDnSWOnwFairM5/Dpa2BcbUhkuEtaM+HhsoCynw4X4WszFokPcri PJYGyF7tPRx5G0qY8TlfsyHKWq570/8RJ8NFa4= Received: (qmail 114479 invoked by alias); 14 Oct 2019 20:57:20 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 114364 invoked by uid 89); 14 Oct 2019 20:57:19 -0000 Authentication-Results: sourceware.org; auth=none X-Spam-SWARE-Status: No, score=-21.2 required=5.0 tests=AWL, BAYES_00, GIT_PATCH_0, GIT_PATCH_1, GIT_PATCH_2, GIT_PATCH_3, KAM_SHORT, RCVD_IN_DNSWL_NONE, SPF_PASS autolearn=ham version=3.3.1 spammy=marks, accomplish X-HELO: mail-qk1-f196.google.com DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:subject:date:message-id:in-reply-to:references; bh=pyXldy314OGfNCQW6cs26lCwSB8A1xaHLx/dTbphUQk=; b=ob+GoW4QhfwC0xFtyRXQdC4cmV9Cu5Ym48Go/e2gn+gi9lcmMXUXb0D/Us9jh/K+x+ byfLbdMfD+9ZQU8FHCEeTk6++XQw1I1g4c+gi9V3mMxZlc6fmfc6ohEZnv3AR9N9WZyq oj/UPY4A7DwrkDi+qLD/ltvnKISMDipskEK5FUcnTJTpDto8086G5FOrbScbmldpsm2k bFaAvpSPIMWWDpbCh07Uz8eZioLnz1Dh9sKUzxNjE0BdnwWaHTTLKUPkUO4Q5aX25AOR 0CmaMv53Jgy3NrK05KRfgQiNZ2KpEXw2jkwvDr4hXDPd2MoQYwBRYIIdLUjmrUYe0Ogg Ovvg== From: Adhemerval Zanella To: libc-alpha@sourceware.org Subject: [PATCH v3 07/21] nptl: i386: Fix Race conditions in pthread cancellation (BZ#12683) Date: Mon, 14 Oct 2019 17:56:42 -0300 Message-Id: <20191014205656.29834-8-adhemerval.zanella@linaro.org> In-Reply-To: <20191014205656.29834-1-adhemerval.zanella@linaro.org> References: <20191014205656.29834-1-adhemerval.zanella@linaro.org> This patch adds the i386 modifications required for the BZ#12683 fix by adding the arch-specific cancellation syscall bridge. The syscall bridge uses the old int80 instruction because by using the optimized vDSO symbol the resulting PC value for an interrupted syscall points to an adress outside the expected markers in __syscall_cancel_arch. It has been discussed in LKML [1] on how kernel could help userland to accomplish it, but afaik discussion has stalled. Also, sysenter should not be used directly by libc since its calling convention is set by the kernel depending of the underlying x86 chip (check kernel commit 30bfa7b3488bfb1bb75c9f50a5fcac1832970c60). Checked on i686-linux-gnu. [1] https://lkml.org/lkml/2016/3/8/1105 --- sysdeps/unix/sysv/linux/i386/Makefile | 2 +- sysdeps/unix/sysv/linux/i386/syscall_cancel.S | 105 ++++++++++++++++++ 2 files changed, 106 insertions(+), 1 deletion(-) create mode 100644 sysdeps/unix/sysv/linux/i386/syscall_cancel.S diff --git a/sysdeps/unix/sysv/linux/i386/Makefile b/sysdeps/unix/sysv/linux/i386/Makefile index da716e2c1b..7928497a35 100644 --- a/sysdeps/unix/sysv/linux/i386/Makefile +++ b/sysdeps/unix/sysv/linux/i386/Makefile @@ -9,7 +9,7 @@ $(objpfx)tst-bz21269: $(shared-thread-library) endif ifeq ($(subdir),elf) -sysdep-dl-routines += libc-do-syscall +sysdep-rtld_routines += libc-do-syscall sysdep-others += lddlibc4 install-bin += lddlibc4 endif diff --git a/sysdeps/unix/sysv/linux/i386/syscall_cancel.S b/sysdeps/unix/sysv/linux/i386/syscall_cancel.S new file mode 100644 index 0000000000..cc8c43f84d --- /dev/null +++ b/sysdeps/unix/sysv/linux/i386/syscall_cancel.S @@ -0,0 +1,105 @@ +/* Cancellable syscall wrapper. Linux/i686 version. + Copyright (C) 2019 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + 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 + . */ + +#include + +/* long int [eax] __syscall_cancel_arch (int *cancelhandling [SP], + long int nr [SP+4], + long int arg1 [SP+8], + long int arg2 [SP+12], + long int arg3 [SP+16], + long int arg4 [SP+20], + long int arg5 [SP+24], + long int arg6 [SP+28]) */ + +ENTRY (__syscall_cancel_arch) + pushl %ebp + cfi_def_cfa_offset (8) + cfi_offset (ebp, -8) + pushl %edi + cfi_def_cfa_offset (12) + cfi_offset (edi, -12) + pushl %esi + cfi_def_cfa_offset (16) + cfi_offset (esi, -16) + pushl %ebx + cfi_def_cfa_offset (20) + cfi_offset (ebx, -20) + + .global __syscall_cancel_arch_start + .type __syscall_cancel_arch_start, @function +__syscall_cancel_arch_start: + + /* if (*cancelhandling & CANCELED_BITMASK) + __syscall_do_cancel() */ + testb $4, (%eax) + jne 1f + + /* Issue a 6 argument syscall, the nr [%eax] being the syscall + number. */ + movl 24(%esp), %eax + movl 28(%esp), %ebx + movl 32(%esp), %ecx + movl 36(%esp), %edx + movl 40(%esp), %esi + movl 44(%esp), %edi + movl 48(%esp), %ebp + + /* We can not use the vDSO helper for syscall (__kernel_vsyscall) + because the returned PC from kernel will point to the vDSO page + instead of the expected __syscall_cancel_arch_{start,end} + marks. */ + int $128 + + .global __syscall_cancel_arch_end + .type __syscall_cancel_arch_end, @function +__syscall_cancel_arch_end: + + popl %ebx + cfi_restore (ebx) + cfi_def_cfa_offset (16) + popl %esi + cfi_restore (esi) + cfi_def_cfa_offset (12) + popl %edi + cfi_restore (edi) + cfi_def_cfa_offset (8) + popl %ebp + cfi_restore (ebp) + cfi_def_cfa_offset (4) + ret + +1: + /* Although the __syscall_do_cancel do not return, we need to stack + being set correctly for unwind. */ + popl %ebx + cfi_restore (ebx) + cfi_def_cfa_offset (16) + popl %esi + cfi_restore (esi) + cfi_def_cfa_offset (12) + popl %edi + cfi_restore (edi) + cfi_def_cfa_offset (8) + popl %ebp + cfi_restore (ebp) + cfi_def_cfa_offset (4) + jmp __syscall_do_cancel + +END (__syscall_cancel_arch) +libc_hidden_def (__syscall_cancel_arch)