From: Christina Schimpe <christina.schimpe@intel.com>
To: gdb-patches@sourceware.org
Cc: thiago.bauermann@linaro.org, eliz@gnu.org
Subject: [PATCH v4 07/11] gdb: Handle shadow stack pointer register unwinding for amd64 linux.
Date: Tue, 17 Jun 2025 12:11:43 +0000 [thread overview]
Message-ID: <20250617121147.1956686-8-christina.schimpe@intel.com> (raw)
In-Reply-To: <20250617121147.1956686-1-christina.schimpe@intel.com>
Unwind the $pl3_ssp register.
We now have an updated value for the shadow stack pointer when
moving up or down the frame level. Note that $pl3_ssp can
become unavailable when moving to a frame before the shadow
stack enablement. In the example below, shadow stack is enabled
in the function 'call1'. Thus, when moving to a frame level above
the function, $pl3_ssp will become unavaiable.
Following the restriction of the linux kernel, implement the unwinding
for amd64 linux only.
Before this patch:
~~~
Breakpoint 1, call2 (j=3) at sample.c:44
44 return 42;
(gdb) p $pl3_ssp
$1 = (void *) 0x7ffff79ffff8
(gdb) up
55 call2 (3);
(gdb) p $pl3_ssp
$2 = (void *) 0x7ffff79ffff8
(gdb) up
68 call1 (43);
(gdb) p $pl3_ssp
$3 = (void *) 0x7ffff79ffff8
~~~
After this patch:
~~~
Breakpoint 1, call2 (j=3) at sample.c:44
44 return 42;
(gdb) p $pl3_ssp
$1 = (void *) 0x7ffff79ffff8
(gdb) up
55 call2 (3);
(gdb) p $pl3_ssp
$2 = (void *) 0x7ffff7a00000
(gdb) up
68 call1 (43i);
(gdb) p $pl3_ssp
$3 = <unavailable>
~~~
As we now have an updated value for each selected frame, the
return command is now enabled for shadow stack enabled programs, too.
We therefore add a test for the return command and shadow stack support,
and for an updated shadow stack pointer after a frame level change.
Reviewed-by: Thiago Jung Bauermann <thiago.bauermann@linaro.org>
---
gdb/amd64-linux-tdep.c | 84 ++++++++++++++++++
gdb/linux-tdep.c | 47 ++++++++++
gdb/linux-tdep.h | 7 ++
.../gdb.arch/amd64-shadow-stack-cmds.exp | 88 +++++++++++++++++++
gdb/testsuite/gdb.arch/amd64-shadow-stack.c | 13 +++
5 files changed, 239 insertions(+)
create mode 100644 gdb/testsuite/gdb.arch/amd64-shadow-stack-cmds.exp
diff --git a/gdb/amd64-linux-tdep.c b/gdb/amd64-linux-tdep.c
index d806d3cb1f7..9436f0b190c 100644
--- a/gdb/amd64-linux-tdep.c
+++ b/gdb/amd64-linux-tdep.c
@@ -47,6 +47,8 @@
#include "arch/amd64-linux-tdesc.h"
#include "inferior.h"
#include "x86-tdep.h"
+#include "dwarf2/frame.h"
+#include "frame-unwind.h"
/* The syscall's XML filename for i386. */
#define XML_SYSCALL_FILENAME_AMD64 "syscalls/amd64-linux.xml"
@@ -1917,6 +1919,87 @@ amd64_linux_get_tls_dtv_addr (struct gdbarch *gdbarch, ptid_t ptid,
return dtv_addr;
}
+/* Return the number of bytes required to update the shadow stack pointer
+ by one element. For x32 the shadow stack elements are still 64-bit
+ aligned. Thus, gdbarch_addr_bit cannot be used to compute the new
+ stack pointer. */
+
+static inline int
+amd64_linux_shadow_stack_element_size_aligned (gdbarch *gdbarch)
+{
+ const bfd_arch_info *binfo = gdbarch_bfd_arch_info (gdbarch);
+ return (binfo->bits_per_word / binfo->bits_per_byte);
+}
+
+
+/* Implement shadow stack pointer unwinding. For each new shadow stack
+ pointer check if its address is still in the shadow stack memory range.
+ If it's outside the range set the returned value to unavailable,
+ otherwise return a value containing the new shadow stack pointer. */
+
+static value *
+amd64_linux_dwarf2_prev_ssp (const frame_info_ptr &this_frame,
+ void **this_cache, int regnum)
+{
+ value *v = frame_unwind_got_register (this_frame, regnum, regnum);
+ gdb_assert (v != nullptr);
+
+ gdbarch *gdbarch = get_frame_arch (this_frame);
+
+ if (v->entirely_available () && !v->optimized_out ())
+ {
+ int size = register_size (gdbarch, regnum);
+ bfd_endian byte_order = gdbarch_byte_order (gdbarch);
+ CORE_ADDR ssp = extract_unsigned_integer (v->contents_all ().data (),
+ size, byte_order);
+
+ /* Starting with v6.6., the Linux kernel supports CET shadow stack.
+ Using /proc/PID/smaps we can only check if the current shadow
+ stack pointer SSP points to shadow stack memory. Only if this is
+ the case a valid previous shadow stack pointer can be
+ calculated. */
+ std::pair<CORE_ADDR, CORE_ADDR> range;
+ if (linux_address_in_shadow_stack_mem_range (ssp, &range))
+ {
+ /* The shadow stack grows downwards. To compute the previous
+ shadow stack pointer, we need to increment SSP. */
+ CORE_ADDR new_ssp
+ = ssp + amd64_linux_shadow_stack_element_size_aligned (gdbarch);
+
+ /* If NEW_SSP points to the end of or before (<=) the current
+ shadow stack memory range we consider NEW_SSP as valid (but
+ empty). */
+ if (new_ssp <= range.second)
+ return frame_unwind_got_address (this_frame, regnum, new_ssp);
+ }
+ }
+
+ /* Return a value which is marked as unavailable in case we could not
+ calculate a valid previous shadow stack pointer. */
+ value *retval
+ = value::allocate_register (get_next_frame_sentinel_okay (this_frame),
+ regnum, register_type (gdbarch, regnum));
+ retval->mark_bytes_unavailable (0, retval->type ()->length ());
+ return retval;
+}
+
+/* Implement the "init_reg" dwarf2_frame_ops method. */
+
+static void
+amd64_init_reg (gdbarch *gdbarch, int regnum, dwarf2_frame_state_reg *reg,
+ const frame_info_ptr &this_frame)
+{
+ if (regnum == gdbarch_pc_regnum (gdbarch))
+ reg->how = DWARF2_FRAME_REG_RA;
+ else if (regnum == gdbarch_sp_regnum (gdbarch))
+ reg->how = DWARF2_FRAME_REG_CFA;
+ else if (regnum == AMD64_PL3_SSP_REGNUM)
+ {
+ reg->how = DWARF2_FRAME_REG_FN;
+ reg->loc.fn = amd64_linux_dwarf2_prev_ssp;
+ }
+}
+
static void
amd64_linux_init_abi_common(struct gdbarch_info info, struct gdbarch *gdbarch,
int num_disp_step_buffers)
@@ -1974,6 +2057,7 @@ amd64_linux_init_abi_common(struct gdbarch_info info, struct gdbarch *gdbarch,
set_gdbarch_remove_non_address_bits_watchpoint
(gdbarch, amd64_linux_remove_non_address_bits_watchpoint);
+ dwarf2_frame_set_init_reg (gdbarch, amd64_init_reg);
}
static void
diff --git a/gdb/linux-tdep.c b/gdb/linux-tdep.c
index 1e339b59e2e..c532b8758bf 100644
--- a/gdb/linux-tdep.c
+++ b/gdb/linux-tdep.c
@@ -47,6 +47,7 @@
#include "gdbsupport/unordered_map.h"
#include <ctype.h>
+#include <algorithm>
/* This enum represents the values that the user can choose when
informing the Linux kernel about which memory mappings will be
@@ -96,6 +97,10 @@ struct smaps_vmflags
/* Memory map has memory tagging enabled. */
unsigned int memory_tagging : 1;
+
+ /* Memory map used for shadow stack. */
+
+ unsigned int shadow_stack_memory : 1;
};
/* Data structure that holds the information contained in the
@@ -537,6 +542,8 @@ decode_vmflags (char *p, struct smaps_vmflags *v)
v->shared_mapping = 1;
else if (strcmp (s, "mt") == 0)
v->memory_tagging = 1;
+ else if (strcmp (s, "ss") == 0)
+ v->shadow_stack_memory = 1;
}
}
@@ -3036,6 +3043,46 @@ show_dump_excluded_mappings (struct ui_file *file, int from_tty,
" flag is %s.\n"), value);
}
+/* See linux-tdep.h. */
+
+bool
+linux_address_in_shadow_stack_mem_range
+ (CORE_ADDR addr, std::pair<CORE_ADDR, CORE_ADDR> *range)
+{
+ if (!target_has_execution () || current_inferior ()->fake_pid_p)
+ return false;
+
+ const int pid = current_inferior ()->pid;
+
+ std::string smaps_file = string_printf ("/proc/%d/smaps", pid);
+
+ gdb::unique_xmalloc_ptr<char> data
+ = target_fileio_read_stralloc (nullptr, smaps_file.c_str ());
+
+ if (data == nullptr)
+ return false;
+
+ const std::vector<smaps_data> smaps
+ = parse_smaps_data (data.get (), std::move (smaps_file));
+
+ auto find_addr_mem_range = [&addr] (const smaps_data &map)
+ {
+ bool addr_in_mem_range
+ = (addr >= map.start_address && addr < map.end_address);
+ return (addr_in_mem_range && map.vmflags.shadow_stack_memory);
+ };
+ auto it = std::find_if (smaps.begin (), smaps.end (), find_addr_mem_range);
+
+ if (it != smaps.end ())
+ {
+ range->first = it->start_address;
+ range->second = it->end_address;
+ return true;
+ }
+
+ return false;
+}
+
/* To be called from the various GDB_OSABI_LINUX handlers for the
various GNU/Linux architectures and machine types.
diff --git a/gdb/linux-tdep.h b/gdb/linux-tdep.h
index 7485fc132a6..7083635b976 100644
--- a/gdb/linux-tdep.h
+++ b/gdb/linux-tdep.h
@@ -117,4 +117,11 @@ extern CORE_ADDR linux_get_hwcap2 ();
extern struct link_map_offsets *linux_ilp32_fetch_link_map_offsets ();
extern struct link_map_offsets *linux_lp64_fetch_link_map_offsets ();
+/* Returns true if ADDR belongs to a shadow stack memory range. If this
+ is the case, assign the shadow stack memory range to RANGE
+ [start_address, end_address). */
+
+extern bool linux_address_in_shadow_stack_mem_range
+ (CORE_ADDR addr, std::pair<CORE_ADDR, CORE_ADDR> *range);
+
#endif /* GDB_LINUX_TDEP_H */
diff --git a/gdb/testsuite/gdb.arch/amd64-shadow-stack-cmds.exp b/gdb/testsuite/gdb.arch/amd64-shadow-stack-cmds.exp
new file mode 100644
index 00000000000..17f32ce3964
--- /dev/null
+++ b/gdb/testsuite/gdb.arch/amd64-shadow-stack-cmds.exp
@@ -0,0 +1,88 @@
+# Copyright 2018-2024 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 <http://www.gnu.org/licenses/>.
+
+# Test shadow stack enabling for frame level update and the return command.
+
+require allow_ssp_tests
+
+standard_testfile amd64-shadow-stack.c
+
+save_vars { ::env(GLIBC_TUNABLES) } {
+
+ append_environment GLIBC_TUNABLES "glibc.cpu.hwcaps" "SHSTK"
+
+ if { [prepare_for_testing "failed to prepare" ${testfile} ${srcfile} \
+ {debug additional_flags="-fcf-protection=return"}] } {
+ return -1
+ }
+
+ clean_restart ${binfile}
+ if { ![runto_main] } {
+ return -1
+ }
+
+ set call1_line [ gdb_get_line_number "break call1" ]
+ set call2_line [ gdb_get_line_number "break call2" ]
+
+ # Extract shadow stack pointer inside main, call1 and call2 function.
+ gdb_breakpoint $call1_line
+ gdb_breakpoint $call2_line
+ set ssp_main [get_valueof /x "\$pl3_ssp" 0 "get value of ssp in main"]
+ gdb_continue_to_breakpoint "break call1" ".*break call1.*"
+ set ssp_call1 [get_valueof /x "\$pl3_ssp" 0 "get value of ssp in call1"]
+ gdb_continue_to_breakpoint "break call2" ".*break call2.*"
+ set ssp_call2 [get_valueof /x "\$pl3_ssp" 0 "get value of ssp in call2"]
+
+ with_test_prefix "test frame level update" {
+ gdb_test "up" "call1.*" "move to frame 1"
+ gdb_test "print /x \$pl3_ssp" "= $ssp_call1" "check pl3_ssp of frame 1"
+ gdb_test "up" "main.*" "move to frame 2"
+ gdb_test "print /x \$pl3_ssp" "= $ssp_main" "check pl3_ssp of frame 2"
+ gdb_test "frame 0" "call2.*" "move to frame 0"
+ gdb_test "print /x \$pl3_ssp" "= $ssp_call2" "check pl3_ssp of frame 0"
+ }
+
+ with_test_prefix "test return from current frame" {
+ gdb_test "return (int) 1" "#0.*call1.*" \
+ "Test shadow stack return from current frame" \
+ "Make.*return now\\? \\(y or n\\) " "y"
+
+ # Potential CET violations often only occur after resuming normal execution.
+ # Therefore, it is important to test normal program continuation after
+ # testing the return command.
+ gdb_continue_to_end
+ }
+
+ clean_restart ${binfile}
+ if { ![runto_main] } {
+ return -1
+ }
+
+ with_test_prefix "test return from past frame" {
+ gdb_breakpoint $call2_line
+ gdb_continue_to_breakpoint "break call2" ".*break call2.*"
+
+ gdb_test "frame 1" ".*in call1.*"
+
+ gdb_test "return (int) 1" "#0.*main.*" \
+ "Test shadow stack return from past frame" \
+ "Make.*return now\\? \\(y or n\\) " "y"
+
+ # Potential CET violations often only occur after resuming normal execution.
+ # Therefore, it is important to test normal program continuation after
+ # testing the return command.
+ gdb_continue_to_end
+ }
+}
diff --git a/gdb/testsuite/gdb.arch/amd64-shadow-stack.c b/gdb/testsuite/gdb.arch/amd64-shadow-stack.c
index 643ef2d5f56..80389730bc2 100644
--- a/gdb/testsuite/gdb.arch/amd64-shadow-stack.c
+++ b/gdb/testsuite/gdb.arch/amd64-shadow-stack.c
@@ -15,8 +15,21 @@
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>. */
+static int
+call2 ()
+{
+ return 42; /* break call2. */
+}
+
+static int
+call1 ()
+{
+ return call2 (); /* break call1. */
+}
+
int
main ()
{
+ call1 (); /* break main. */
return 0;
}
--
2.34.1
Intel Deutschland GmbH
Registered Address: Am Campeon 10, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de
Managing Directors: Sean Fennelly, Jeffrey Schneiderman, Tiffany Doon Silva
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928
next prev parent reply other threads:[~2025-06-17 12:16 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-06-17 12:11 [PATCH v4 00/11] Add CET shadow stack support Christina Schimpe
2025-06-17 12:11 ` [PATCH v4 01/11] gdbserver: Add optional runtime register set type Christina Schimpe
2025-06-19 9:27 ` Luis Machado
2025-06-17 12:11 ` [PATCH v4 02/11] gdbserver: Add assert in x86_linux_read_description Christina Schimpe
2025-06-19 9:27 ` Luis Machado
2025-06-17 12:11 ` [PATCH v4 03/11] gdb: Sync up x86-gcc-cpuid.h with cpuid.h from gcc 14 branch Christina Schimpe
2025-06-17 18:12 ` Tom Tromey
2025-06-20 12:39 ` Schimpe, Christina
2025-06-17 12:11 ` [PATCH v4 04/11] gdb, gdbserver: Use xstate_bv for target description creation on x86 Christina Schimpe
2025-06-19 9:23 ` Luis Machado
2025-06-23 12:46 ` Schimpe, Christina
2025-06-23 12:56 ` Luis Machado
2025-06-24 13:46 ` Schimpe, Christina
2025-06-26 16:03 ` Luis Machado
2025-06-17 12:11 ` [PATCH v4 05/11] gdb, gdbserver: Add support of Intel shadow stack pointer register Christina Schimpe
2025-06-17 12:20 ` Eli Zaretskii
2025-06-19 9:24 ` Luis Machado
2025-06-23 13:05 ` Schimpe, Christina
2025-06-17 12:11 ` [PATCH v4 06/11] gdb: amd64 linux coredump support with shadow stack Christina Schimpe
2025-06-19 9:24 ` Luis Machado
2025-06-23 13:16 ` Schimpe, Christina
2025-06-17 12:11 ` Christina Schimpe [this message]
2025-06-19 9:25 ` [PATCH v4 07/11] gdb: Handle shadow stack pointer register unwinding for amd64 linux Luis Machado
2025-06-20 1:42 ` Thiago Jung Bauermann
2025-06-23 14:55 ` Schimpe, Christina
2025-06-23 23:26 ` Thiago Jung Bauermann
2025-06-23 15:00 ` Schimpe, Christina
2025-06-23 15:06 ` Luis Machado
2025-06-23 23:36 ` Thiago Jung Bauermann
2025-06-20 1:52 ` Thiago Jung Bauermann
2025-06-17 12:11 ` [PATCH v4 08/11] gdb, gdbarch: Enable inferior calls for shadow stack support Christina Schimpe
2025-06-19 9:25 ` Luis Machado
2025-06-23 17:49 ` Schimpe, Christina
2025-06-17 12:11 ` [PATCH v4 09/11] gdb: Implement amd64 linux shadow stack support for inferior calls Christina Schimpe
2025-06-17 12:21 ` Eli Zaretskii
2025-06-19 9:25 ` Luis Machado
2025-06-27 19:52 ` Schimpe, Christina
2025-06-28 10:38 ` Luis Machado
2025-06-28 20:03 ` Thiago Jung Bauermann
2025-06-28 21:05 ` Thiago Jung Bauermann
2025-06-17 12:11 ` [PATCH v4 10/11] gdb, gdbarch: Introduce gdbarch method to get the shadow stack pointer Christina Schimpe
2025-06-17 18:16 ` Tom Tromey
2025-06-20 12:59 ` Schimpe, Christina
2025-06-19 9:26 ` Luis Machado
2025-06-23 18:00 ` Schimpe, Christina
2025-06-17 12:11 ` [PATCH v4 11/11] gdb: Enable displaced stepping with shadow stack on amd64 linux Christina Schimpe
2025-06-17 12:22 ` Eli Zaretskii
2025-06-17 15:16 ` Schimpe, Christina
2025-06-19 9:26 ` Luis Machado
2025-06-23 18:24 ` Schimpe, Christina
2025-06-24 8:05 ` Luis Machado
2025-06-27 19:26 ` Schimpe, Christina
2025-06-28 10:35 ` Luis Machado
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250617121147.1956686-8-christina.schimpe@intel.com \
--to=christina.schimpe@intel.com \
--cc=eliz@gnu.org \
--cc=gdb-patches@sourceware.org \
--cc=thiago.bauermann@linaro.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox