diff options
author | Andrew Burgess <andrew.burgess@embecosm.com> | 2021-03-02 14:19:56 +0000 |
---|---|---|
committer | Andrew Burgess <andrew.burgess@embecosm.com> | 2021-03-26 17:43:14 +0000 |
commit | b1f3973b9c55a1406c13ade5cb14239d5b17a465 (patch) | |
tree | 39fbe51feb95f05ae16b74796e617163d8616200 /bfd/ChangeLog | |
parent | b4b1a226df8d51da9594200ad803ad303c15cd36 (diff) | |
download | gdb-b1f3973b9c55a1406c13ade5cb14239d5b17a465.zip gdb-b1f3973b9c55a1406c13ade5cb14239d5b17a465.tar.gz gdb-b1f3973b9c55a1406c13ade5cb14239d5b17a465.tar.bz2 |
gdb/testsuite: more testing of pretty printer 'array' display_hint
This commit adds a couple of tests to the python pretty printer
testing.
I've added a test for the 'array' display hint. This display hint is
tested by gdb.python/py-mi.exp, however, the MI testing is done via
the varobj interface, and this code makes its own direct calls to the
Python pretty printers from gdb/varobj.c. What this means is that the
interface to the pretty printers in gdb/python/py-prettyprint.c is not
tested for the 'array' display hint path.
I also added a test for what happens when the display_hint method
raises an exception. There wasn't a bug that inspired this test, just
while adding the previous test I thought, I wonder what happens if...
The current behaviour of GDB seems reasonable, GDB displays the Python
exception, and then continues printing the value as if display_hint
had returned None. I added a test to lock in this behaviour.
gdb/testsuite/ChangeLog:
* gdb.python/py-prettyprint.c (struct container): Add 'is_array_p'
member.
(make_container): Initialise is_array_p.
* gdb.python/py-prettyprint.exp: Add new tests.
* gdb.python/py-prettyprint.py (ContainerPrinter.display_hint):
Check is_array_p and possibly return 'array'.
Diffstat (limited to 'bfd/ChangeLog')
0 files changed, 0 insertions, 0 deletions