New /s modifier for the disassemble command.

The "source centric" /m option to the disassemble command is often
unhelpful, e.g., in the presence of optimized code.
This patch adds a /s modifier that is better.
For one, /m only prints instructions from the originating source file,
leaving out instructions from e.g., inlined functions from other files.

gdb/ChangeLog:

	PR gdb/11833
	* NEWS: Document new /s modifier for the disassemble command.
	* cli/cli-cmds.c (disassemble_command): Add support for /s.
	(_initialize_cli_cmds): Update online docs of disassemble command.
	* disasm.c: #include "source.h".
	(struct deprecated_dis_line_entry): Renamed from dis_line_entry.
	All uses updated.
	(dis_line_entry): New struct.
	(hash_dis_line_entry, eq_dis_line_entry): New functions.
	(allocate_dis_line_table): New functions.
	(maybe_add_dis_line_entry, line_has_code_p): New functions.
	(dump_insns): New arg end_pc.  All callers updated.
	(do_mixed_source_and_assembly_deprecated): Renamed from
	do_mixed_source_and_assembly.  All callers updated.
	(do_mixed_source_and_assembly): New function.
	(gdb_disassembly): Handle /s (DISASSEMBLY_SOURCE).
	* disasm.h (DISASSEMBLY_SOURCE_DEPRECATED): Renamed from
	DISASSEMBLY_SOURCE.  All uses updated.
	(DISASSEMBLY_SOURCE): New macro.
	* mi/mi-cmd-disas.c (mi_cmd_disassemble): New modes 4,5.

gdb/doc/ChangeLog:

	* gdb.texinfo (Machine Code): Update docs for mixed source/assembly
	disassembly.
	(GDB/MI Data Manipulation): Update docs for new disassembly modes.

gdb/testsuite/ChangeLog:

	* gdb.mi/mi-disassemble.exp: Update.
	* gdb.base/disasm-optim.S: New file.
	* gdb.base/disasm-optim.c: New file.
	* gdb.base/disasm-optim.h: New file.
	* gdb.base/disasm-optim.exp: New file.

2 files changed, 118 insertions, 9 deletions

diff --git a/gdb/doc/ChangeLog b/gdb/doc/ChangeLog
index 5b3d9a1..806dac4 100644
--- a/gdb/doc/ChangeLog
+++ b/gdb/doc/ChangeLog
@@ -1,3 +1,9 @@
+2015-08-14  Doug Evans  <xdje42@gmail.com>
+
+	* gdb.texinfo (Machine Code): Update docs for mixed source/assembly
+	disassembly.
+	(GDB/MI Data Manipulation): Update docs for new disassembly modes.
+
 2015-08-11  Keith Seitz  <keiths@redhat.com>
 
 	* gdb.texinfo (Thread-Specific Breakpoints, Printing Source Lines):
diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index c64c7a1..c876ed6 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -8098,11 +8098,12 @@ Variables}).
 @cindex listing machine instructions
 @item disassemble
 @itemx disassemble /m
+@itemx disassemble /s
 @itemx disassemble /r
 This specialized command dumps a range of memory as machine
 instructions.  It can also print mixed source+disassembly by specifying
-the @code{/m} modifier and print the raw instructions in hex as well as
-in symbolic form by specifying the @code{/r}.
+the @code{/m} or @code{/s} modifier and print the raw instructions in hex
+as well as in symbolic form by specifying the @code{/r} modifier.
 The default memory range is the function surrounding the
 program counter of the selected frame.  A single argument to this
 command is a program counter value; @value{GDBN} dumps the function
@@ -8146,8 +8147,9 @@ Dump of assembler code from 0x32c4 to 0x32e4:
 End of assembler dump.
 @end smallexample
 
-Here is an example showing mixed source+assembly for Intel x86, when the
-program is stopped just after function prologue:
+Here is an example showing mixed source+assembly for Intel x86
+with @code{/m} or @code{/s}, when the program is stopped just after
+function prologue in a non-optimized function with no inline code.
 
 @smallexample
 (@value{GDBP}) disas /m main
@@ -8172,6 +8174,96 @@ Dump of assembler code for function main:
 End of assembler dump.
 @end smallexample
 
+The @code{/m} option is deprecated as its output is not useful when
+there is either inlined code or re-ordered code.
+The @code{/s} option is the preferred choice.
+Here is an example for AMD x86-64 showing the difference between
+@code{/m} output and @code{/s} output.
+This example has one inline function defined in a header file,
+and the code is compiled with @samp{-O2} optimization.
+Note how the @code{/m} output is missing the disassembly of
+several instructions that are present in the @code{/s} output.
+
+@file{foo.h}:
+
+@smallexample
+int
+foo (int a)
+@{
+  if (a < 0)
+    return a * 2;
+  if (a == 0)
+    return 1;
+  return a + 10;
+@}
+@end smallexample
+
+@file{foo.c}:
+
+@smallexample
+#include "foo.h"
+volatile int x, y;
+int
+main ()
+@{
+  x = foo (y);
+  return 0;
+@}
+@end smallexample
+
+@smallexample
+(@value{GDBP}) disas /m main
+Dump of assembler code for function main:
+5	@{
+
+6	  x = foo (y);
+   0x0000000000400400 <+0>:	mov    0x200c2e(%rip),%eax # 0x601034 <y>
+   0x0000000000400417 <+23>:	mov    %eax,0x200c13(%rip) # 0x601030 <x>
+
+7	  return 0;
+8	@}
+   0x000000000040041d <+29>:	xor    %eax,%eax
+   0x000000000040041f <+31>:	retq
+   0x0000000000400420 <+32>:	add    %eax,%eax
+   0x0000000000400422 <+34>:	jmp    0x400417 <main+23>
+
+End of assembler dump.
+(@value{GDBP}) disas /s main
+Dump of assembler code for function main:
+foo.c:
+5	@{
+6	  x = foo (y);
+   0x0000000000400400 <+0>:	mov    0x200c2e(%rip),%eax # 0x601034 <y>
+
+foo.h:
+4	  if (a < 0)
+   0x0000000000400406 <+6>:	test   %eax,%eax
+   0x0000000000400408 <+8>:	js     0x400420 <main+32>
+
+6	  if (a == 0)
+7	    return 1;
+8	  return a + 10;
+   0x000000000040040a <+10>:	lea    0xa(%rax),%edx
+   0x000000000040040d <+13>:	test   %eax,%eax
+   0x000000000040040f <+15>:	mov    $0x1,%eax
+   0x0000000000400414 <+20>:	cmovne %edx,%eax
+
+foo.c:
+6	  x = foo (y);
+   0x0000000000400417 <+23>:	mov    %eax,0x200c13(%rip) # 0x601030 <x>
+
+7	  return 0;
+8	@}
+   0x000000000040041d <+29>:	xor    %eax,%eax
+   0x000000000040041f <+31>:	retq
+
+foo.h:
+5	    return a * 2;
+   0x0000000000400420 <+32>:	add    %eax,%eax
+   0x0000000000400422 <+34>:	jmp    0x400417 <main+23>
+End of assembler dump.
+@end smallexample
+
 Here is another example showing raw instructions in hex for AMD x86-64,
 
 @smallexample
@@ -29782,9 +29874,20 @@ displayed; if @var{lines} is higher than the number of lines between
 @var{start-addr} and @var{end-addr}, only the lines up to @var{end-addr}
 are displayed.
 @item @var{mode}
-is either 0 (meaning only disassembly), 1 (meaning mixed source and
-disassembly), 2 (meaning disassembly with raw opcodes), or 3 (meaning
-mixed source and disassembly with raw opcodes).
+is one of:
+@itemize @bullet
+@item 0 disassembly only
+@item 1 mixed source and disassembly (deprecated)
+@item 2 disassembly with raw opcodes
+@item 3 mixed source and disassembly with raw opcodes (deprecated)
+@item 4 mixed source and disassembly
+@item 5 mixed source and disassembly with raw opcodes
+@end itemize
+
+Modes 1 and 3 are deprecated.  The output is ``source centric''
+which hasn't proved useful in practice.
+@xref{Machine Code}, for a discussion of the difference between
+@code{/m} and @code{/s} output of the @code{disassemble} command.
 @end table
 
 @subsubheading Result
@@ -29810,12 +29913,12 @@ The decimal offset in bytes from the start of @samp{func-name}.
 The text disassembly for this @samp{address}.
 
 @item opcodes
-This field is only present for mode 2.  This contains the raw opcode
+This field is only present for modes 2, 3 and 5.  This contains the raw opcode
 bytes for the @samp{inst} field.
 
 @end table
 
-For modes 1 and 3 the @samp{asm_insns} list contains tuples named
+For modes 1, 3, 4 and 5 the @samp{asm_insns} list contains tuples named
 @samp{src_and_asm_line}, each of which has the following fields:
 
 @table @code