From patchwork Wed Mar 27 07:47:38 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Schimpe, Christina" X-Patchwork-Id: 87687 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 B2B1B3858286 for ; Wed, 27 Mar 2024 07:48:43 +0000 (GMT) X-Original-To: gdb-patches@sourceware.org Delivered-To: gdb-patches@sourceware.org Received: from mgamail.intel.com (mgamail.intel.com [192.198.163.17]) by sourceware.org (Postfix) with ESMTPS id 92E05385E453 for ; Wed, 27 Mar 2024 07:48:06 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 92E05385E453 Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=intel.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=intel.com ARC-Filter: OpenARC Filter v1.0.0 sourceware.org 92E05385E453 Authentication-Results: server2.sourceware.org; arc=none smtp.remote-ip=192.198.163.17 ARC-Seal: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1711525689; cv=none; b=Sg4FlDbelTHAeEnwstDBfZr+EihHwZ0KMjcBEHHehA4oCuU11F2XuH1TFmaUpoH6qlLsA1gST9tucIjj8qQO5Uyh4p3tT3RWFPRmXF+h1UWldRWeC3w9ajaOO7k6xIEhYKk2Z043553UanZNbiy6ACkuc98UXa23znkAhhuhSRc= ARC-Message-Signature: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1711525689; c=relaxed/simple; bh=HNEPWyS67nWN2CBUnXIa9wBwvV+mJnZn1oKXZi0enac=; h=DKIM-Signature:From:To:Subject:Date:Message-Id:MIME-Version; b=E2uoAmG3v6pF6t3oIJkNs/TUUPC+/fQmHJW2q53YyhPCkwepFJJXYMG6rxGF6Kn2MNMv0Cwd77/clGtFdTqfxR3NBQCQf0Ilbg7GuHoq6r/2LJavbSh97BhZxQeEa6LzklWc3NCw0TgyWfkh9ZeeFrB4e+HPeKCbQBF7HCg2lew= ARC-Authentication-Results: i=1; server2.sourceware.org DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1711525687; x=1743061687; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=HNEPWyS67nWN2CBUnXIa9wBwvV+mJnZn1oKXZi0enac=; b=R/B5W/mhdp0EZqC4B+NemQi6bP8vHhPtOq7eTvofv6TUOj8mC2RtK2a5 y/9uMu8+Bbe3Qn2yAYE7qt5r1LdqLkH0xik54jPHl9X4UEkqwy6HATu83 b/YORMNYyuhWuZG96vTGAsGQqw1x5ewsOsU/x+kt9oq3gMqhjJfTUEnC5 vflYI33lxicOF8xJmZQXNU73XIXqhPFaeaLH6cnlT8EZMr3+IHzwqp5dG H2EQgs3Bl8omxdftnD5kqNQ+ppnT2z2LWW4Y1j0AGPIZBVAXLpKo3pKTi 2w17LQlmp7fno3m45Tuwerl+AA13tnvGPBDiwVYwfLfQakC730wkaHjuE g==; X-CSE-ConnectionGUID: gqkndwROQYy1EpfZxcgtvw== X-CSE-MsgGUID: 0q2g86mYSG6fPvgZcHyeSg== X-IronPort-AV: E=McAfee;i="6600,9927,11025"; a="6473665" X-IronPort-AV: E=Sophos;i="6.07,158,1708416000"; d="scan'208";a="6473665" Received: from fmviesa002.fm.intel.com ([10.60.135.142]) by fmvoesa111.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 27 Mar 2024 00:48:06 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.07,158,1708416000"; d="scan'208";a="39324662" Received: from gkldtt-dev-004.igk.intel.com (HELO localhost) ([10.123.221.202]) by fmviesa002-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 27 Mar 2024 00:48:05 -0700 From: "Schimpe, Christina" To: gdb-patches@sourceware.org Cc: christina.schimpe@intel.com Subject: [PATCH 2/3] LAM: Enable tagged pointer support for watchpoints. Date: Wed, 27 Mar 2024 07:47:38 +0000 Message-Id: <20240327074739.2969623-3-christina.schimpe@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20240327074739.2969623-1-christina.schimpe@intel.com> References: <20240327074739.2969623-1-christina.schimpe@intel.com> MIME-Version: 1.0 X-Spam-Status: No, score=-10.1 required=5.0 tests=BAYES_00, DKIMWL_WL_HIGH, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, GIT_PATCH_0, KAM_SHORT, RCVD_IN_DNSWL_NONE, SPF_HELO_NONE, SPF_NONE, TXREP 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: gdb-patches@sourceware.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Gdb-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: gdb-patches-bounces+patchwork=sourceware.org@sourceware.org From: Christina Schimpe The Intel (R) linear address masking (LAM) feature modifies the checking applied to 64-bit linear addresses. With this so-called "modified canonicality check" the processor masks the metadata bits in a pointer before using it as a linear address. LAM supports two different modes that differ regarding which pointer bits are masked and can be used for metadata: LAM 48 resulting in a LAM width of 15 and LAM 57 resulting in a LAM width of 6. This patch adjusts watchpoint addresses based on the currently enabled LAM mode using the untag mask provided in the /proc//status file. As LAM can be enabled at runtime or as the configuration may change when entering an enclave, GDB checks enablement state each time a watchpoint is updated. In contrast to the patch implemented for ARM's Top Byte Ignore "Clear non-significant bits of address on memory access", it is not necessary to adjust addresses before they are passed to the target layer cache, as for LAM tagged pointers are supported by the system call to read memory. Additionally, LAM applies only to addresses used for data accesses. Thus, it is sufficient to mask addresses used for watchpoints. The following examples are based on a LAM57 enabled program. Before this patch tagged pointers were not supported for watchpoints: ~~~ (gdb) print pi_tagged $2 = (int *) 0x10007ffffffffe004 (gdb) watch *pi_tagged Hardware watchpoint 2: *pi_tagged (gdb) c Continuing. Couldn't write debug register: Invalid argument. ~~~~ Once LAM 48 or LAM 57 is enabled for the current program, GDB can now specify watchpoints for tagged addresses with LAM width 15 or 6, respectively. Reviewed-By: Eli Zaretskii --- gdb/NEWS | 2 + gdb/amd64-linux-tdep.c | 67 ++++++++++++++++++++++++++++ gdb/testsuite/gdb.arch/amd64-lam.c | 49 ++++++++++++++++++++ gdb/testsuite/gdb.arch/amd64-lam.exp | 45 +++++++++++++++++++ gdb/testsuite/lib/gdb.exp | 62 +++++++++++++++++++++++++ 5 files changed, 225 insertions(+) create mode 100755 gdb/testsuite/gdb.arch/amd64-lam.c create mode 100644 gdb/testsuite/gdb.arch/amd64-lam.exp diff --git a/gdb/NEWS b/gdb/NEWS index feb3a37393a..295f2147def 100644 --- a/gdb/NEWS +++ b/gdb/NEWS @@ -3,6 +3,8 @@ *** Changes since GDB 14 +* GDB now supports watchpoints for tagged data pointers on amd64. + * The MPX commands "show/set mpx bound" have been deprecated, as Intel listed MPX as removed in 2019. diff --git a/gdb/amd64-linux-tdep.c b/gdb/amd64-linux-tdep.c index 9d560ac4fbf..94c62277623 100644 --- a/gdb/amd64-linux-tdep.c +++ b/gdb/amd64-linux-tdep.c @@ -41,6 +41,7 @@ #include "arch/amd64.h" #include "target-descriptions.h" #include "expop.h" +#include "inferior.h" /* The syscall's XML filename for i386. */ #define XML_SYSCALL_FILENAME_AMD64 "syscalls/amd64-linux.xml" @@ -48,6 +49,10 @@ #include "record-full.h" #include "linux-record.h" +#include + +#define DEFAULT_TAG_MASK 0xffffffffffffffffULL + /* Mapping between the general-purpose registers in `struct user' format and GDB's register cache layout. */ @@ -1794,6 +1799,65 @@ amd64_dtrace_parse_probe_argument (struct gdbarch *gdbarch, } } +/* Extract the untagging mask based on the currently active linear address + masking (LAM) mode, which is stored in the /proc//status file. + If we cannot extract the untag mask (for example, if we don't have + execution), we assume address tagging is not enabled and return the + DEFAULT_TAG_MASK. */ + +static CORE_ADDR +amd64_linux_lam_untag_mask () +{ + if (!target_has_execution ()) + return DEFAULT_TAG_MASK; + + inferior *inf = current_inferior (); + if (inf->fake_pid_p) + return DEFAULT_TAG_MASK; + + /* Construct status file name and read the file's content. */ + std::string filename = string_printf ("/proc/%d/status", inf->pid); + gdb::unique_xmalloc_ptr status_file + = target_fileio_read_stralloc (nullptr, filename.c_str ()); + + if (status_file == nullptr) + return DEFAULT_TAG_MASK; + + /* Parse the status file line-by-line and look for the untag mask. */ + std::istringstream strm_status_file (status_file.get ()); + std::string line; + const std::string untag_mask_str ("untag_mask:\t"); + while (std::getline (strm_status_file, line)) + { + const size_t found = line.find (untag_mask_str); + if (found != std::string::npos) + { + const size_t tag_length = untag_mask_str.length(); + return std::strtoul (&line[found + tag_length], nullptr, 0); + } + } + + return DEFAULT_TAG_MASK; +} + +/* Adjust watchpoint address based on the tagging mode which is currently + enabled. For now, linear address masking (LAM) is the only feature + which allows to store metadata in pointer values for amd64. Thus, we + adjust the watchpoint address based on the currently active LAM mode + using the untag mask provided by the linux kernel. Check each time for + a new mask, as LAM is enabled at runtime. Also, the LAM configuration + may change when entering an enclave. No untagging will be applied if + this function is called while we don't have execution. */ + +static CORE_ADDR +amd64_linux_remove_non_addr_bits_wpt (gdbarch *gdbarch, CORE_ADDR addr) +{ + /* Clear insignificant bits of a target address using the untag mask. + The untag mask preserves the topmost bit, which distinguishes user space + from kernel space address. */ + return (addr & amd64_linux_lam_untag_mask ()); +} + static void amd64_linux_init_abi_common(struct gdbarch_info info, struct gdbarch *gdbarch, int num_disp_step_buffers) @@ -1848,6 +1912,9 @@ amd64_linux_init_abi_common(struct gdbarch_info info, struct gdbarch *gdbarch, set_gdbarch_get_siginfo_type (gdbarch, x86_linux_get_siginfo_type); set_gdbarch_report_signal_info (gdbarch, i386_linux_report_signal_info); + + set_gdbarch_remove_non_addr_bits_wpt (gdbarch, + amd64_linux_remove_non_addr_bits_wpt); } static void diff --git a/gdb/testsuite/gdb.arch/amd64-lam.c b/gdb/testsuite/gdb.arch/amd64-lam.c new file mode 100755 index 00000000000..28786389a9a --- /dev/null +++ b/gdb/testsuite/gdb.arch/amd64-lam.c @@ -0,0 +1,49 @@ +/* This testcase is part of GDB, the GNU debugger. + + Copyright 2023 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 3 of the License, or + (at your option) any later version. + + This program 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 General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . */ + +#define _GNU_SOURCE +#include +#include +#include +#include +#include +#include +#include + +int +main (int argc, char **argv) +{ + int i; + int* pi = &i; + int* pi_tagged; + + /* Enable LAM 57. */ + errno = 0; + syscall (SYS_arch_prctl, ARCH_ENABLE_TAGGED_ADDR, 6); + assert_perror (errno); + + /* Add tagging at bit 61. */ + pi_tagged = (int *) ((uintptr_t) pi | (1LL << 60)); + + i = 0; /* Breakpoint here. */ + *pi = 1; + *pi_tagged = 2; + *pi = 3; + *pi_tagged = 4; + + return 0; +} diff --git a/gdb/testsuite/gdb.arch/amd64-lam.exp b/gdb/testsuite/gdb.arch/amd64-lam.exp new file mode 100644 index 00000000000..8274c3adf97 --- /dev/null +++ b/gdb/testsuite/gdb.arch/amd64-lam.exp @@ -0,0 +1,45 @@ +# Copyright 2023 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program 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 General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see . + +# Test Linear Address Masking (LAM) support. + +require allow_lam_tests + +standard_testfile amd64-lam.c + +# Test LAM 57. +if { [prepare_for_testing "failed to prepare" ${testfile} ${srcfile}] } { + return -1 +} + +if { ![runto_main] } { + return -1 +} + +gdb_breakpoint [gdb_get_line_number "Breakpoint here"] +gdb_continue_to_breakpoint "Breakpoint here" + +# Test hw watchpoint for tagged and untagged address with hit on untagged +# and tagged adress. +foreach symbol {"pi" "pi_tagged"} { + gdb_test "watch *${symbol}" + gdb_test "continue" \ + "Continuing\\..*Hardware watchpoint \[0-9\]+.*" \ + "run until watchpoint on ${symbol}" + gdb_test "continue" \ + "Continuing\\..*Hardware watchpoint \[0-9\]+.*" \ + "run until watchpoint on ${symbol}, 2nd hit" + delete_breakpoints +} diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp index d48ea37c0cc..c7d15f82da1 100644 --- a/gdb/testsuite/lib/gdb.exp +++ b/gdb/testsuite/lib/gdb.exp @@ -9285,6 +9285,68 @@ gdb_caching_proc allow_ctf_tests {} { return $can_ctf } +# Run a test on the target to see if it supports LAM 57. Return 1 if so, +# 0 if it does not. Based on the arch_prctl() handle ARCH_ENABLE_TAGGED_ADDR +# to enable LAM which fails if the hardware or the OS does not support LAM. + +gdb_caching_proc allow_lam_tests {} { + global gdb_prompt inferior_exited_re + + set me "allow_lam_tests" + if { ![istarget "x86_64-*-*"] } { + verbose "$me: target does not support LAM, returning 1" 2 + return 0 + } + + # Compile a test program. + set src { + #define _GNU_SOURCE + #include + #include + #include + #include + + int configure_lam () + { + errno = 0; + syscall (SYS_arch_prctl, ARCH_ENABLE_TAGGED_ADDR, 6); + assert_perror (errno); + return errno; + } + + int + main () { return configure_lam (); } + } + + if {![gdb_simple_compile $me $src executable ""]} { + return 0 + } + # No error message, compilation succeeded so now run it via gdb. + + set allow_lam_tests 0 + clean_restart $obj + gdb_run_cmd + gdb_expect { + -re ".*$inferior_exited_re with code.*${gdb_prompt} $" { + verbose -log "$me: LAM support not detected." + } + -re ".*Program received signal SIGABRT, Aborted.*${gdb_prompt} $" { + verbose -log "$me: LAM support not detected." + } + -re ".*$inferior_exited_re normally.*${gdb_prompt} $" { + verbose -log "$me: LAM support detected." + set allow_lam_tests 1 + } + default { + warning "\n$me: default case taken." + } + } + gdb_exit + remote_file build delete $obj + + verbose "$me: returning $allow_lam_tests" 2 + return $allow_lam_tests +} # Return 1 if compiler supports -gstatement-frontiers. Otherwise, # return 0.