* [PATCH][gdb] Fix segv in "maint print symbols" for ada exec
@ 2020-03-10 14:08 Tom de Vries
2020-03-10 16:40 ` Tom Tromey
0 siblings, 1 reply; 2+ messages in thread
From: Tom de Vries @ 2020-03-10 14:08 UTC (permalink / raw)
To: gdb-patches
Hi,
When using the executable from test-case gdb.ada/access_to_packed_array.exp
(read-in using -readnow) and printing the symbols using "maint print symbols",
we run into a segv:
...
$ gdb -readnow -batch access_to_packed_array/foo -ex "maint print symbols"
...
info: array (<>) of character; computed at runtime
ptr: range 0 .. 2147483647; computed at runtime
Aborted (core dumped)
...
What happens is that dwarf2_evaluate_property gets called and sets the local
frame variable to the current frame, which happens to be NULL. Subsequently
the PROP_LOCLIST handling code is executed, where get_frame_address_in_block
gets called with argument NULL, and the segv is triggered.
Fix this by handling a NULL frame in the PROP_LOCLIST handling code in
dwarf2_evaluate_property.
Build and reg-tested on x86_64-linux.
OK for trunk?
Thanks,
- Tom
[gdb] Fix segv in "maint print symbols" for ada exec
gdb/ChangeLog:
2020-03-10 Tom de Vries <tdevries@suse.de>
* dwarf2/loc.c (dwarf2_evaluate_property): Handle NULL frame in
PROP_LOCLIST handling code.
gdb/testsuite/ChangeLog:
2020-03-10 Tom de Vries <tdevries@suse.de>
* gdb.ada/access_to_packed_array.exp: Test printing of expanded
symtabs.
---
gdb/dwarf2/loc.c | 6 +++++-
gdb/testsuite/gdb.ada/access_to_packed_array.exp | 6 ++++++
2 files changed, 11 insertions(+), 1 deletion(-)
diff --git a/gdb/dwarf2/loc.c b/gdb/dwarf2/loc.c
index a5074da8bf..5155cff60d 100644
--- a/gdb/dwarf2/loc.c
+++ b/gdb/dwarf2/loc.c
@@ -2511,11 +2511,15 @@ dwarf2_evaluate_property (const struct dynamic_prop *prop,
{
struct dwarf2_property_baton *baton
= (struct dwarf2_property_baton *) prop->data.baton;
- CORE_ADDR pc = get_frame_address_in_block (frame);
+ CORE_ADDR pc;
const gdb_byte *data;
struct value *val;
size_t size;
+ if (frame == NULL
+ || !get_frame_address_in_block_if_available (frame, &pc))
+ return false;
+
data = dwarf2_find_location_expression (&baton->loclist, &size, pc);
if (data != NULL)
{
diff --git a/gdb/testsuite/gdb.ada/access_to_packed_array.exp b/gdb/testsuite/gdb.ada/access_to_packed_array.exp
index 61ad230ec0..a2a80a1659 100644
--- a/gdb/testsuite/gdb.ada/access_to_packed_array.exp
+++ b/gdb/testsuite/gdb.ada/access_to_packed_array.exp
@@ -22,6 +22,12 @@ if {[gdb_compile_ada "${srcfile}" "${binfile}" executable [list debug]] != "" }
}
clean_restart ${testfile}
+gdb_test_no_output "maint expand-symtabs"
+gdb_test "set logging redirect on"
+gdb_test "set logging on"
+gdb_test "maint print symbols"
+gdb_test "set logging off"
+file delete gdb.txt
set bp_location [gdb_get_line_number "BREAK" ${testdir}/foo.adb]
runto "foo.adb:$bp_location"
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: [PATCH][gdb] Fix segv in "maint print symbols" for ada exec
2020-03-10 14:08 [PATCH][gdb] Fix segv in "maint print symbols" for ada exec Tom de Vries
@ 2020-03-10 16:40 ` Tom Tromey
0 siblings, 0 replies; 2+ messages in thread
From: Tom Tromey @ 2020-03-10 16:40 UTC (permalink / raw)
To: Tom de Vries; +Cc: gdb-patches
>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
Tom> What happens is that dwarf2_evaluate_property gets called and sets the local
Tom> frame variable to the current frame, which happens to be NULL. Subsequently
Tom> the PROP_LOCLIST handling code is executed, where get_frame_address_in_block
Tom> gets called with argument NULL, and the segv is triggered.
Tom> Fix this by handling a NULL frame in the PROP_LOCLIST handling code in
Tom> dwarf2_evaluate_property.
Thank you for doing this. This seems reasonable.
I had one request...
Tom> clean_restart ${testfile}
Tom> +gdb_test_no_output "maint expand-symtabs"
Tom> +gdb_test "set logging redirect on"
Tom> +gdb_test "set logging on"
Tom> +gdb_test "maint print symbols"
Tom> +gdb_test "set logging off"
Tom> +file delete gdb.txt
I think this code should set the logging file to something in the
output directory. This way if other tests do the same thing, there
won't be a possibility of a clash.
After this, there's no real need to delete the file; though it's fine if
you still want to do that.
Tom
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-03-10 16:40 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-03-10 14:08 [PATCH][gdb] Fix segv in "maint print symbols" for ada exec Tom de Vries
2020-03-10 16:40 ` Tom Tromey
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox