[PATCH] Add instruction bytes to DAP disassembly response

[PATCH] Add instruction bytes to DAP disassembly response

[Tom Tromey]

The DAP disassemble command lets the client return the underlying
bytes of the instruction in an implementation-defined format.  This
patch updates gdb to return this, and simply uses a hex string of the
bytes as the format.
---
 gdb/doc/gdb.texinfo                   | 5 +++++
 gdb/python/lib/gdb/dap/disassemble.py | 5 ++++-
 gdb/testsuite/gdb.dap/basic-dap.exp   | 9 +++++++++
 3 files changed, 18 insertions(+), 1 deletion(-)

diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index b10c06ae91f..714f4bc1672 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -39087,6 +39087,11 @@ The target to which @value{GDBN} should connect.  This is a string and
 is passed to the @code{target remote} command.  @xref{Connecting}.
 @end table
 
+In response to the @code{disassemble} request, DAP allows the client
+to return the bytes of each instruction in an implementation-defined
+format.  @value{GDBN} implements this by sending a string with the
+bytes encoded in hex, like @code{"55a2b900"}.
+
 @node JIT Interface
 @chapter JIT Compilation Interface
 @cindex just-in-time compilation
diff --git a/gdb/python/lib/gdb/dap/disassemble.py b/gdb/python/lib/gdb/dap/disassemble.py
index bc091eb2c89..dda2f43b5a3 100644
--- a/gdb/python/lib/gdb/dap/disassemble.py
+++ b/gdb/python/lib/gdb/dap/disassemble.py
@@ -21,18 +21,21 @@ from .startup import send_gdb_with_response, in_gdb_thread
 
 @in_gdb_thread
 def _disassemble(pc, skip_insns, count):
+    inf = gdb.selected_inferior()
     try:
         arch = gdb.selected_frame().architecture()
     except gdb.error:
         # Maybe there was no frame.
-        arch = gdb.selected_inferior().architecture()
+        arch = inf.architecture()
     result = []
     total_count = skip_insns + count
     for elt in arch.disassemble(pc, count=total_count)[skip_insns:]:
+        mem = inf.read_memory(elt["addr"], elt["length"])
         result.append(
             {
                 "address": hex(elt["addr"]),
                 "instruction": elt["asm"],
+                "instructionBytes": mem.hex(),
             }
         )
     return {
diff --git a/gdb/testsuite/gdb.dap/basic-dap.exp b/gdb/testsuite/gdb.dap/basic-dap.exp
index df9afcfcdfc..81c63a4ac09 100644
--- a/gdb/testsuite/gdb.dap/basic-dap.exp
+++ b/gdb/testsuite/gdb.dap/basic-dap.exp
@@ -191,6 +191,15 @@ set obj [dap_check_request_and_response "disassemble one instruction" \
 		  $insn_pc]]
 set response [lindex $obj 0]
 gdb_assert { [dict exists $response body instructions] } "instructions in disassemble output"
+foreach insn [dict get $response body instructions] {
+    gdb_assert {[dict exists $insn instructionBytes]} \
+	"instruction bytes in disassemble output"
+    set bytes [dict get $insn instructionBytes]
+    gdb_assert {[string length $bytes] % 2 == 0} \
+	"even number of digits"
+    gdb_assert {[regexp "^\[0-9A-Fa-f\]+\$" $bytes]} \
+	"instructionBytes is hex"
+}
 
 set obj [dap_check_request_and_response "command repl" \
 	     evaluate {o expression [s {output 23}] context [s repl]}]
-- 
2.40.1

Re: [PATCH] Add instruction bytes to DAP disassembly response

[Eli Zaretskii]

> Cc: Tom Tromey <tromey@adacore.com>
> Date: Mon,  3 Jul 2023 07:59:02 -0600
> From: Tom Tromey via Gdb-patches <gdb-patches@sourceware.org>
> 
> The DAP disassemble command lets the client return the underlying
> bytes of the instruction in an implementation-defined format.  This
> patch updates gdb to return this, and simply uses a hex string of the
> bytes as the format.
> ---
>  gdb/doc/gdb.texinfo                   | 5 +++++
>  gdb/python/lib/gdb/dap/disassemble.py | 5 ++++-
>  gdb/testsuite/gdb.dap/basic-dap.exp   | 9 +++++++++
>  3 files changed, 18 insertions(+), 1 deletion(-)

Thanks, the documentation part is OK.

Reviewed-By: Eli Zaretskii <eliz@gnu.org>

Re: [PATCH] Add instruction bytes to DAP disassembly response

[Tom Tromey]

>>>>> "Tom" == Tom Tromey <tromey@adacore.com> writes:

Tom> The DAP disassemble command lets the client return the underlying
Tom> bytes of the instruction in an implementation-defined format.  This
Tom> patch updates gdb to return this, and simply uses a hex string of the
Tom> bytes as the format.

I'm checking this in.

Tom