This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [PREC/RFA] Add not_replay to make precord support release memory better

From: Michael Snyder <msnyder at vmware dot com>
To: Hui Zhu <teawater at gmail dot com>
Cc: gdb-patches ml <gdb-patches at sourceware dot org>
Date: Thu, 23 Jul 2009 19:10:53 -0700
Subject: Re: [PREC/RFA] Add not_replay to make precord support release memory better
References: <daef60380907210122t35962b01nf0a393407061a67b@mail.gmail.com>

Hui Zhu wrote:

Hi,

Now, in replay reverse mode, if get error with memory entry, it will
output error.  But it just affect the operation with this part of
memory.
So I make a patch to let it set the not_replay in memory entry to 1 in
this status.  When replay forward mode, this memory entry will not
really set to memory.


I like the idea.  I like it a lot!
Can this solve the problem with sbrk() too?
(code comments below)

For example:
#include <sys/types.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <errno.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <stdint.h>
#include <sys/mman.h>

int
main(int argc,char *argv[],char *envp[])
{
	char	*buf;

	buf = mmap (0, 1024, PROT_READ | PROT_WRITE, MAP_PRIVATE |
MAP_ANONYMOUS, -1, 0);
	if (buf == (caddr_t) -1) {
		perror ("mmap");
		return (-errno);
	}

buf[0] = 1;

munmap (buf, 1024);

	return (0);
}

Before patch:
gdb ~/gdb/a.out
GNU gdb (GDB) 6.8.50.20090720-cvs
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Setting up the environment for debugging gdb.
Function "internal_error" not defined.
Make breakpoint pending on future shared library load? (y or [n])
[answered N; input not from terminal]
Function "info_command" not defined.
Make breakpoint pending on future shared library load? (y or [n])
[answered N; input not from terminal]
/home/teawater/gdb/bgdbno/gdb/.gdbinit:8: Error in sourced command file:
No breakpoint number 0.
(gdb) start
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
Temporary breakpoint 1 at 0x8048425: file 1.c, line 17.
Starting program: /home/teawater/gdb/a.out

Temporary breakpoint 1, main (argc=<value optimized out>, argv=<value
optimized out>, envp=<value optimized out>)
    at 1.c:17
17		buf = mmap (0, 1024, PROT_READ | PROT_WRITE, MAP_PRIVATE |
MAP_ANONYMOUS, -1, 0);
(gdb) record
(gdb) n
During symbol reading, incomplete CFI data; unspecified registers
(e.g., eax) at 0x8048422.
18		if (buf == (caddr_t) -1) {
(gdb)
23		buf[0] = 1;
(gdb)
25		munmap (buf, 1024);
(gdb)
The next instruction is syscall munmap.  It will free the memory addr
= 0xb7fe0000 len = 1024.  It will make record target get error.  Do
you want to stop the program?([y] or n) n
27		return (0);
(gdb) rc
Continuing.
Process record: error reading memory at addr = 0xb7fe0000 len = 1.
(gdb)


After patch:
./gdb ~/gdb/a.out
GNU gdb (GDB) 6.8.50.20090721-cvs
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Setting up the environment for debugging gdb.
Function "internal_error" not defined.
Make breakpoint pending on future shared library load? (y or [n])
[answered N; input not from terminal]
Function "info_command" not defined.
Make breakpoint pending on future shared library load? (y or [n])
[answered N; input not from terminal]
/home/teawater/gdb/bgdbno/gdb/.gdbinit:8: Error in sourced command file:
No breakpoint number 0.
(gdb) start
During symbol reading, DW_AT_name missing from DW_TAG_base_type.
Temporary breakpoint 1 at 0x8048425: file 1.c, line 17.
Starting program: /home/teawater/gdb/a.out

Temporary breakpoint 1, main (argc=<value optimized out>, argv=<value
optimized out>, envp=<value optimized out>)
    at 1.c:17
17		buf = mmap (0, 1024, PROT_READ | PROT_WRITE, MAP_PRIVATE |
MAP_ANONYMOUS, -1, 0);
(gdb) record
(gdb) n
During symbol reading, incomplete CFI data; unspecified registers
(e.g., eax) at 0x8048422.
18		if (buf == (caddr_t) -1) {
(gdb)
23		buf[0] = 1;
(gdb)
25		munmap (buf, 1024);
(gdb)
The next instruction is syscall munmap.  It will free the memory addr
= 0xb7fe0000 len = 1024.  It will make record target get error.  Do
you want to stop the program?([y] or n) n
27		return (0);
(gdb) rc
Continuing.

No more reverse-execution history.
main (argc=<value optimized out>, argv=<value optimized out>,
envp=<value optimized out>) at 1.c:17
17		buf = mmap (0, 1024, PROT_READ | PROT_WRITE, MAP_PRIVATE |
MAP_ANONYMOUS, -1, 0);
(gdb) n
18		if (buf == (caddr_t) -1) {
(gdb)
23		buf[0] = 1;
(gdb)
25		munmap (buf, 1024);
(gdb) rn
23		buf[0] = 1;

Please help me review it.

Thanks,
Hui

2009-07-21  Hui Zhu  <teawater@gmail.com>
	
	* record.c (record_mem_entry): Add not_replay.  If this the
	memory entry doesn't replay, it will set to 1.
	(record_arch_list_add_mem): Initialize not_replay to 0.
	(record_wait): In replay reverse mode, if get error with
	memory entry, not output error, just set not_replay to 1.
	In replay forward mode, if not_replay is set, don't set
	memory entry to memory.


A changelog entry doesn't need to explain so much.
This much explanation belongs in comments in the code.

Change log entries explain more "what was changed",
and not so much "why".

This should say something like:

	* record.c (record_mem_entry): New field 'not_replay'.
	(record_arch_list_add_mem): Initialize not_replay to 0.
	(record_wait): Set 'not_replay' flag if target memory
	not readable.  Don't try to change target memory if
	'not_replay' is set.

---
 record.c |   75 ++++++++++++++++++++++++++++++++++++++++++---------------------
 1 file changed, 50 insertions(+), 25 deletions(-)

--- a/record.c
+++ b/record.c
@@ -51,6 +51,7 @@ struct record_mem_entry
 {
   CORE_ADDR addr;
   int len;
+  int not_replay;

This needs a more descriptive name (and some comments!)

How about something like:

    /* Set this flag if target memory for this entry
       can no longer be accessed.  */
    int mem_entry_not_accessible;

   gdb_byte *val;
 };

@@ -275,6 +276,7 @@ record_arch_list_add_mem (CORE_ADDR addr
   rec->type = record_mem;
   rec->u.mem.addr = addr;
   rec->u.mem.len = len;
+  rec->u.mem.not_replay = 0;

   if (target_read_memory (addr, rec->u.mem.val, len))
     {
@@ -727,32 +729,55 @@ record_wait (struct target_ops *ops,
 	  else if (record_list->type == record_mem)
 	    {
 	      /* mem */
-	      gdb_byte *mem = alloca (record_list->u.mem.len);
-	      if (record_debug > 1)
-		fprintf_unfiltered (gdb_stdlog,
-				    "Process record: record_mem %s to "
-				    "inferior addr = %s len = %d.\n",
-				    host_address_to_string (record_list),
-				    paddress (gdbarch, record_list->u.mem.addr),
-				    record_list->u.mem.len);
-
-	      if (target_read_memory
-		  (record_list->u.mem.addr, mem, record_list->u.mem.len))
-		error (_("Process record: error reading memory at "
-			 "addr = %s len = %d."),
-		       paddress (gdbarch, record_list->u.mem.addr),
-		       record_list->u.mem.len);
-
-	      if (target_write_memory
-		  (record_list->u.mem.addr, record_list->u.mem.val,
-		   record_list->u.mem.len))
-		error (_
-		       ("Process record: error writing memory at "
-			"addr = %s len = %d."),
-		       paddress (gdbarch, record_list->u.mem.addr),
-		       record_list->u.mem.len);
+	      if (record_list->u.mem.not_replay
+		  && execution_direction != EXEC_REVERSE)
+		record_list->u.mem.not_replay = 0;

First of all, a comment to explain what's going on, please.

Now -- if I understand correctly (and please tell if I'm wrong),

1) During the recording "pass", there's no change.
2) During the reverse-replay pass, if the memory is
not readable or writable, we will set this flag to TRUE.
3) Finally, during the forward-replay pass, if the flag
has previously been set to TRUE, we will skip this entry
(and set the flag to FALSE.)

So my question is -- is there any reason to set it to FALSE here?
Is there any way that the memory can ever become readable again?

Seems to me, once it is set TRUE, we might as well just leave it TRUE.
Am I right?

+	      else
+		{
+		  gdb_byte *mem = alloca (record_list->u.mem.len);
+		  if (record_debug > 1)
+		    fprintf_unfiltered (gdb_stdlog,
+				        "Process record: record_mem %s to "
+				        "inferior addr = %s len = %d.\n",
+				        host_address_to_string (record_list),
+				        paddress (gdbarch,
+					          record_list->u.mem.addr),
+				        record_list->u.mem.len);

-	      memcpy (record_list->u.mem.val, mem, record_list->u.mem.len);
+		  if (target_read_memory (record_list->u.mem.addr, mem,
+		                          record_list->u.mem.len))
+	            {
+		      if (execution_direction != EXEC_REVERSE)
+		        error (_("Process record: error reading memory at "
+			         "addr = %s len = %d."),
+		               paddress (gdbarch, record_list->u.mem.addr),
+		               record_list->u.mem.len);
+		      else
+		        record_list->u.mem.not_replay = 1;

A comment to explain what's happening, please.

+		    }
+		  else
+		    {
+		      if (target_write_memory (record_list->u.mem.addr,
+			                       record_list->u.mem.val,
+		                               record_list->u.mem.len))
+	                {
+			  if (execution_direction != EXEC_REVERSE)
+			    error (_("Process record: error writing memory at "
+			             "addr = %s len = %d."),
+		                   paddress (gdbarch, record_list->u.mem.addr),
+		                   record_list->u.mem.len);
+			  else
+			    {
+			      record_list->u.mem.not_replay = 1;


And again, a comment.  This whole section needs more comments,
but let's start with this, before it becomes even more difficult
to understand.

+			    }
+			}
+		      else
+		        {
+			  memcpy (record_list->u.mem.val, mem,
+				  record_list->u.mem.len);
+			}
+		    }
+		}
 	    }
 	  else
 	    {

Follow-Ups:
- Re: [PREC/RFA] Add not_replay to make precord support release memory better
  - From: Hui Zhu

References:
- [PREC/RFA] Add not_replay to make precord support release memory better
  - From: Hui Zhu

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]