diff mbox

[pushed] gdb/python: Return None from Progspace.block_for_pc on error

Message ID 20191024152506.67E65238BA@gnutoolchain-gerrit.osci.io
State New
Headers show

Commit Message

Simon Marchi (Code Review) Oct. 24, 2019, 3:25 p.m. UTC
The original change was created by Andrew Burgess.

Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/223
......................................................................

gdb/python: Return None from Progspace.block_for_pc on error

The documentation for Progspace.block_for_pc says:

  Return the innermost gdb.Block containing the given pc value. If the
  block cannot be found for the pc value specified, the function will
  return None.

However, the implementation actually throws an error for invalid
addresses, like this:

    (gdb) python print gdb.current_progspace ().block_for_pc (1)
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
    RuntimeError: Cannot locate object file for block.
    Error while executing Python code.
    (gdb)

This has been the behaviour since the command was first added (when
the documentation was still as above) in this commit:

    commit f3e9a8177c41893858fce2bdf339dbe90b3a4ef5
    Date:   Wed Feb 24 21:18:28 2010 +0000

Since that commit the code in question has moved around, but the
important parts are largely unchanged.  The function in question is
now in py-progspace.c:pspy_block_for_pc.

Examining the code shows that the real state is more complex than just
the function throws an error instead of returning None, instead the
real situation is:

  1. If we can't find a compilation unit for the $pc value then we
  throw an error, but

  2. If we can find a compilation unit, but can't find a block within
  the compilation unit for the $pc then return None.

I suspect for most users of the Python API this distinction is
irrelevant, and I propose that we standardise on one single failure
mechanism.

Given the function can currently return None in some cases, and is
documented to return None on error, I propose we make that the case
for all error paths, which is what this patch does.

As the Progspace.block_for_pc method is currently untested, I've added
some basic tests including for a call with an invalid $pc.

This is potentially an API breaking change, though an undocumented
part of the API.  Also, users should have been checking and handling a
None return value anyway, so my hope is that this shouldn't be too
disruptive.

gdb/ChangeLog:

	* python/py-progspace.c (pspy_block_for_pc): Return None for all
	error paths.

gdb/testsuite/ChangeLog:

	* gdb.python/py-progspace.exp: Add tests for the
	Progspace.block_for_pc method.

Change-Id: I9cea8d2132902bcad0013d1fd39080dd5423cc57
---
M gdb/ChangeLog
M gdb/python/py-progspace.c
M gdb/testsuite/ChangeLog
M gdb/testsuite/gdb.python/py-progspace.exp
4 files changed, 25 insertions(+), 5 deletions(-)
diff mbox

Patch

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index 09adb91..887c7fb 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,3 +1,8 @@ 
+2019-10-24  Andrew Burgess  <andrew.burgess@embecosm.com>
+
+	* python/py-progspace.c (pspy_block_for_pc): Return None for all
+	error paths.
+
 2019-10-23  Tom Tromey  <tom@tromey.com>
 
 	* arc-tdep.c: Remove ".." from include.
diff --git a/gdb/python/py-progspace.c b/gdb/python/py-progspace.c
index 4483d03..bdb7072 100644
--- a/gdb/python/py-progspace.c
+++ b/gdb/python/py-progspace.c
@@ -397,11 +397,7 @@ 
     }
 
   if (cust == NULL || COMPUNIT_OBJFILE (cust) == NULL)
-    {
-      PyErr_SetString (PyExc_RuntimeError,
-		       _("Cannot locate object file for block."));
-      return NULL;
-    }
+    Py_RETURN_NONE;
 
   if (block)
     return block_to_block_object (block, COMPUNIT_OBJFILE (cust));
diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
index 0d1fed9..12ef1fe 100644
--- a/gdb/testsuite/ChangeLog
+++ b/gdb/testsuite/ChangeLog
@@ -1,3 +1,8 @@ 
+2019-10-24  Andrew Burgess  <andrew.burgess@embecosm.com>
+
+	* gdb.python/py-progspace.exp: Add tests for the
+	Progspace.block_for_pc method.
+
 2019-10-23  Tom Tromey  <tom@tromey.com>
 
 	* configure: Rebuild.
diff --git a/gdb/testsuite/gdb.python/py-progspace.exp b/gdb/testsuite/gdb.python/py-progspace.exp
index 5394382..d1bcb81 100644
--- a/gdb/testsuite/gdb.python/py-progspace.exp
+++ b/gdb/testsuite/gdb.python/py-progspace.exp
@@ -57,6 +57,20 @@ 
     return
 }
 
+# Check we can get a block for the current $pc.
+set pc_val [get_integer_valueof "\$pc" 0]
+gdb_py_test_silent_cmd "python blk = gdb.current_progspace ().block_for_pc (${pc_val})" \
+    "get block for the current \$pc" 1
+gdb_test "python print blk.start <= ${pc_val}" "True" \
+    "block start is before \$pc"
+gdb_test "python print blk.end >= ${pc_val}" "True" \
+    "block end is after \$pc"
+
+# Check what happens when we ask for a block of an invalid address.
+if ![is_address_zero_readable] {
+    gdb_test "python print gdb.current_progspace ().block_for_pc (0)" "None"
+}
+
 # With a single inferior, progspace.objfiles () and gdb.objfiles () should
 # be identical.
 gdb_test "python print (progspace.objfiles () == gdb.objfiles ())" "True"