public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Stefan Liebler <stli@linux.vnet.ibm.com>
To: libc-alpha@sourceware.org
Subject: Re: [PATCH] S390: Sync ptrace.h with kernel. [BZ #21539]
Date: Mon, 19 Jun 2017 13:11:00 -0000	[thread overview]
Message-ID: <a70194fb-d96c-d22f-7aba-36be9685d421@linux.vnet.ibm.com> (raw)
In-Reply-To: <20170613200522.GA14306@altlinux.org>

[-- Attachment #1: Type: text/plain, Size: 2130 bytes --]

On 06/13/2017 10:05 PM, Dmitry V. Levin wrote:
> On Tue, Jun 06, 2017 at 12:17:33PM +0200, Stefan Liebler wrote:
> [...]
>> diff --git a/sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c b/sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c
>> new file mode 100644
 >> [...]
>> +      /* Get information about tracee: gprs, last breaking address.  */
>> +      parea.len = sizeof (regs);
>> +      parea.process_addr = (unsigned long) &regs;
>> +      parea.kernel_addr = 0;
>> +      ptrace (PTRACE_PEEKUSR_AREA, pid, &parea);
> 
> Note that you can verify whether PTRACE_PEEKUSR_AREA has returned
> the expected result by comparing registers with those returned
> by PTRACE_GETREGSET.  The latter is implemented on s390 since
> linux 2.6.27 so its use in glibc is safe.
> 
Okay. Now the gprs are obtained by PTRACE_PEEKUSR_AREA and 
PTRACE_GETREGSET. Afterwards I use memcmp to check whether the values 
are the same.
>> +      ptrace (PTRACE_GET_LAST_BREAK, pid, NULL, &last_break);
> 
> As these ptrace calls are expected to succeed,
> you might want to check their return code.
> 
Done with several usages of TEST_VERIFY_EXIT.

>> +
>> +      printf ("child IA: %p last_break: %p\n",
>> +	      (void *) regs[1], (void *) last_break);
>> +
>> +      /* Execute tracee until next taken branch.
>> +
>> +	 Note:
>> +	 Before the commit which introduced this testcase,
>> +	 <glibc>/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
>> +	 uses ptrace-request 12 for PTRACE_GETREGS,
>> +	 but <kernel>/include/uapi/linux/ptrace.h
>> +	 uses 12 for PTRACE_SINGLEBLOCK.
>> +
>> +	 The s390 kernel has no support for PTRACE_GETREGS!
>> +	 Thus glibc ptrace.h is adjusted to match kernel ptrace.h.
>> +
>> +	 This test ensures, that PTRACE_SINGLEBLOCK defined in glibc
>> +	 works as expected.  If the kernel would interpret it as
>> +	 PTRACE_GETREGS, then the tracee will not make any progress
>> +	 and this testcase will time out.  */
>> +      ptrace (req_singleblock, pid, NULL, NULL);
> 
> Likewise.
> 
> 


I've attached the patch with the mentioned changes and the NEWS entry 
requested by Andreas.

Is this okay?

Bye
Stefan

[-- Attachment #2: 20170619_s390_ptraceh.patch --]
[-- Type: text/x-patch, Size: 10287 bytes --]

commit 99d6dcd30803d992b2e82c1edb8db797f4ace846
Author: Stefan Liebler <stli@linux.vnet.ibm.com>
Date:   Mon Jun 19 15:02:02 2017 +0200

    S390: Sync ptrace.h with kernel. [BZ #21539]
    
    This patch removes PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS
    and PTRACE_SETFPREGS as these requests does not exist on s390 kernel.
    
    But the kernel has support for PTRACE_SINGLEBLOCK,
    PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA, PTRACE_POKEUSR_AREA,
    PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE, PTRACE_DISABLE_TE and
    PTRACE_TE_ABORT_RAND.  Thus those are defined now.
    
    The current kernel s390 specific ptrace.h file also defines
    PTRACE_PEEKTEXT_AREA, PTRACE_PEEKDATA_AREA, PTRACE_POKETEXT_AREA,
    PTRACE_POKEDATA_AREA, PTRACE_PEEK_SYSTEM_CALL, PTRACE_POKE_SYSTEM_CALL
    and PTRACE_PROT, but those requests are not supported.
    Thus those defines are skipped in glibc ptrace.h.
    
    There were old includes of ptrace.h in sysdeps/s390/fpu/fesetenv.c.
    The ptrace feature isn't used there anymore, thus I removed the includes.
    
    Before this patch, <glibc>/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
    uses ptrace-request 12 for PTRACE_GETREGS,
    but <kernel>/include/uapi/linux/ptrace.h uses 12 for PTRACE_SINGLEBLOCK.
    
    The s390 kernel has never had support for PTRACE_GETREGS!
    Thus glibc ptrace.h is adjusted to match kernel ptrace.h.
    
    The new s390 specific test ensures, that PTRACE_SINGLEBLOCK defined
    in glibc works as expected.  If the kernel would interpret it as
    PTRACE_GETREGS, then the testcase will not make any progress
    and will time out.
    
    ChangeLog:
    
    	[BZ #21539]
    	* NEWS: Mention s390 ptrace request changes.
    	* sysdeps/unix/sysv/linux/s390/sys/ptrace.h
    	(PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS,
    	PTRACE_SETFPREGS): Remove enum constant.
    	(PT_GETREGS, PT_SETREGS, PT_GETFPREGS, T_SETFPREGS):
    	Remove defines.
    	(PTRACE_SINGLEBLOCK): New enum constant.
    	(PT_STEPBLOCK): New define.
    	(PTRACE_PEEKUSR_AREA, PTRACE_POKEUSR_AREA,
    	PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE, PTRACE_DISABLE_TE,
    	PTRACE_TE_ABORT_RAND): New enum constant and define.
    	* sysdeps/s390/fpu/fesetenv.c: Remove ptrace.h includes.
    	* sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c:
    	New file.
    	* sysdeps/unix/sysv/linux/s390/Makefile: Add test.

diff --git a/NEWS b/NEWS
index 804c1b9..9ee2218 100644
--- a/NEWS
+++ b/NEWS
@@ -98,6 +98,15 @@ Version 2.26
 * The tunables feature is now enabled by default.  This allows users to tweak
   behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
 
+* The s390 specific ptrace requests are adjusted to the kernel ones.  Request 12
+  is now used for PTRACE_SINGLEBLOCK instead of PTRACE_GETREGS.  The requests
+  PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS were
+  removed as those are not supported by the s390 kernel.  The requests
+  PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
+  PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
+  PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND were added as those are supported
+  by the s390 kernel.
+
 Security related changes:
 
 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
diff --git a/sysdeps/s390/fpu/fesetenv.c b/sysdeps/s390/fpu/fesetenv.c
index 4c9bcf0..0f64a3f 100644
--- a/sysdeps/s390/fpu/fesetenv.c
+++ b/sysdeps/s390/fpu/fesetenv.c
@@ -20,8 +20,6 @@
 #include <fenv_libc.h>
 #include <fpu_control.h>
 #include <stddef.h>
-#include <asm/ptrace.h>
-#include <sys/ptrace.h>
 #include <unistd.h>
 
 int
diff --git a/sysdeps/unix/sysv/linux/s390/Makefile b/sysdeps/unix/sysv/linux/s390/Makefile
index 3867c33..f30a6bb 100644
--- a/sysdeps/unix/sysv/linux/s390/Makefile
+++ b/sysdeps/unix/sysv/linux/s390/Makefile
@@ -29,3 +29,7 @@ CFLAGS-elision-trylock.c = $(elision-CFLAGS)
 CFLAGS-elision-unlock.c = $(elision-CFLAGS)
 endif
 endif
+
+ifeq ($(subdir),misc)
+tests += tst-ptrace-singleblock
+endif
diff --git a/sysdeps/unix/sysv/linux/s390/sys/ptrace.h b/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
index 7caf101..88079fc 100644
--- a/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
+++ b/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
@@ -89,25 +89,9 @@ enum __ptrace_request
   PTRACE_SINGLESTEP = 9,
 #define PT_STEP PTRACE_SINGLESTEP
 
-  /* Get all general purpose registers used by a processes.
-     This is not supported on all machines.  */
-   PTRACE_GETREGS = 12,
-#define PT_GETREGS PTRACE_GETREGS
-
-  /* Set all general purpose registers used by a processes.
-     This is not supported on all machines.  */
-   PTRACE_SETREGS = 13,
-#define PT_SETREGS PTRACE_SETREGS
-
-  /* Get all floating point registers used by a processes.
-     This is not supported on all machines.  */
-   PTRACE_GETFPREGS = 14,
-#define PT_GETFPREGS PTRACE_GETFPREGS
-
-  /* Set all floating point registers used by a processes.
-     This is not supported on all machines.  */
-   PTRACE_SETFPREGS = 15,
-#define PT_SETFPREGS PTRACE_SETFPREGS
+  /* Execute process until next taken branch.  */
+  PTRACE_SINGLEBLOCK = 12,
+#define PT_STEPBLOCK PTRACE_SINGLEBLOCK
 
   /* Attach to a process that is already running. */
   PTRACE_ATTACH = 16,
@@ -167,8 +151,26 @@ enum __ptrace_request
   PTRACE_SETSIGMASK = 0x420b,
 #define PTRACE_SETSIGMASK PTRACE_SETSIGMASK
 
-  PTRACE_SECCOMP_GET_FILTER = 0x420c
+  PTRACE_SECCOMP_GET_FILTER = 0x420c,
 #define PTRACE_SECCOMP_GET_FILTER PTRACE_SECCOMP_GET_FILTER
+
+  PTRACE_PEEKUSR_AREA = 0x5000,
+#define PTRACE_PEEKUSR_AREA PTRACE_PEEKUSR_AREA
+
+  PTRACE_POKEUSR_AREA = 0x5001,
+#define PTRACE_POKEUSR_AREA PTRACE_POKEUSR_AREA
+
+  PTRACE_GET_LAST_BREAK = 0x5006,
+#define PTRACE_GET_LAST_BREAK PTRACE_GET_LAST_BREAK
+
+  PTRACE_ENABLE_TE = 0x5009,
+#define PTRACE_ENABLE_TE PTRACE_ENABLE_TE
+
+  PTRACE_DISABLE_TE = 0x5010,
+#define PTRACE_DISABLE_TE PTRACE_DISABLE_TE
+
+  PTRACE_TE_ABORT_RAND = 0x5011
+#define PTRACE_TE_ABORT_RAND PTRACE_TE_ABORT_RAND
 };
 
 
diff --git a/sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c b/sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c
new file mode 100644
index 0000000..95a2f55
--- /dev/null
+++ b/sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c
@@ -0,0 +1,128 @@
+/* Testing s390x PTRACE_SINGLEBLOCK ptrace request.
+   Copyright (C) 2017 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
+   <http://www.gnu.org/licenses/>.  */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <sys/wait.h>
+#include <sys/types.h>
+#include <sys/uio.h>
+#include <elf.h>
+#include <support/xunistd.h>
+#include <support/check.h>
+
+/* Ensure that we use the PTRACE_SINGLEBLOCK definition from glibc ptrace.h
+   in tracer_func.  We need the kernel ptrace.h for structs ptrace_area
+   and gregset_t.  */
+#include <sys/ptrace.h>
+static const enum __ptrace_request req_singleblock = PTRACE_SINGLEBLOCK;
+#include <asm/ptrace.h>
+
+static void
+tracee_func (int pid)
+{
+  /* Dump the mapping information for manual inspection of the printed
+     tracee addresses.  */
+  char str[80];
+  sprintf (str, "cat /proc/%d/maps", pid);
+  puts (str);
+  system (str);
+  fflush (stdout);
+
+  TEST_VERIFY_EXIT (ptrace (PTRACE_TRACEME) == 0);
+  /* Stop tracee.  Afterwards the tracer_func can operate.  */
+  kill (pid, SIGSTOP);
+
+  puts ("The PTRACE_SINGLEBLOCK of the tracer will stop after: "
+	"brasl %r14,<puts@plt>!");
+}
+
+static void
+tracer_func (int pid)
+{
+  unsigned long last_break;
+  ptrace_area parea;
+  gregset_t regs;
+  struct iovec parea2;
+  gregset_t regs2;
+
+  int status;
+
+  while (1)
+    {
+      /* Wait for the tracee to be stopped or exited.  */
+      wait (&status);
+      if (WIFEXITED (status))
+	break;
+
+      /* Get information about tracee: gprs, last breaking address.  */
+      parea.len = sizeof (regs);
+      parea.process_addr = (unsigned long) &regs;
+      parea.kernel_addr = 0;
+      TEST_VERIFY_EXIT (ptrace (PTRACE_PEEKUSR_AREA, pid, &parea) == 0);
+      TEST_VERIFY_EXIT (ptrace (PTRACE_GET_LAST_BREAK, pid, NULL, &last_break)
+			== 0);
+
+      parea2.iov_len = sizeof (regs2);
+      parea2.iov_base = &regs2;
+      TEST_VERIFY_EXIT (ptrace (PTRACE_GETREGSET, pid, NT_PRSTATUS, &parea2)
+			== 0);
+      TEST_VERIFY_EXIT (parea2.iov_len == sizeof (regs2));
+
+      /* Test if gprs obtained by PTRACE_PEEKUSR_AREA and PTRACE_GETREGESET
+	 have the same values.  */
+      TEST_VERIFY_EXIT (memcmp (&regs, &regs2, sizeof (regs)) == 0);
+
+      printf ("child IA: %p last_break: %p\n",
+	      (void *) regs[1], (void *) last_break);
+
+      /* Execute tracee until next taken branch.
+
+	 Note:
+	 Before the commit which introduced this testcase,
+	 <glibc>/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
+	 uses ptrace-request 12 for PTRACE_GETREGS,
+	 but <kernel>/include/uapi/linux/ptrace.h
+	 uses 12 for PTRACE_SINGLEBLOCK.
+
+	 The s390 kernel has no support for PTRACE_GETREGS!
+	 Thus glibc ptrace.h is adjusted to match kernel ptrace.h.
+
+	 This test ensures, that PTRACE_SINGLEBLOCK defined in glibc
+	 works as expected.  If the kernel would interpret it as
+	 PTRACE_GETREGS, then the tracee will not make any progress
+	 and this testcase will time out.  */
+      TEST_VERIFY_EXIT (ptrace (req_singleblock, pid, NULL, NULL) == 0);
+    }
+}
+
+static int
+do_test (void)
+{
+  int pid;
+  pid = xfork ();
+  if (pid)
+    tracer_func (pid);
+  else
+    tracee_func (getpid ());
+
+  return EXIT_SUCCESS;
+}
+
+#include <support/test-driver.c>

  reply	other threads:[~2017-06-19 13:11 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-06-06 10:17 Stefan Liebler
2017-06-06 10:44 ` Andreas Schwab
2017-06-06 10:58   ` Dmitry V. Levin
2017-06-06 11:56   ` Stefan Liebler
2017-06-08 12:02     ` Stefan Liebler
2017-06-13 20:05 ` Dmitry V. Levin
2017-06-19 13:11   ` Stefan Liebler [this message]
2017-06-19 13:26     ` Dmitry V. Levin
2017-06-19 14:34       ` Stefan Liebler
2017-06-30 10:09     ` Florian Weimer
2017-07-04  8:22       ` Stefan Liebler
2017-07-04  9:41         ` Florian Weimer
2017-07-04 15:37           ` Stefan Liebler
2017-07-07 10:22             ` Stefan Liebler
2017-07-07 10:45               ` Florian Weimer
2017-07-07 13:54                 ` Stefan Liebler
2017-07-11  8:39                   ` Stefan Liebler
2017-07-18 10:20     ` Dmitry V. Levin
2017-07-18 13:31       ` Carlos O'Donell
2017-07-18 13:39         ` Dmitry V. Levin
2017-07-18 14:11           ` Carlos O'Donell
2017-07-18 14:28             ` Mark Wielaard
2017-07-18 14:40               ` Mark Wielaard
2017-07-20  7:38             ` Stefan Liebler
2017-07-20  8:07               ` Carlos O'Donell
2017-07-20  8:32                 ` Stefan Liebler
2017-07-24  3:51                   ` Dmitry V. Levin
2017-07-24  7:18                     ` Stefan Liebler
2017-07-18 13:41       ` Stefan Liebler
2017-07-18 14:12         ` Carlos O'Donell
2017-07-18 14:16         ` Dmitry V. Levin
2017-07-19  8:40           ` Stefan Liebler

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=a70194fb-d96c-d22f-7aba-36be9685d421@linux.vnet.ibm.com \
    --to=stli@linux.vnet.ibm.com \
    --cc=libc-alpha@sourceware.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).