On Mon, Jun 23, 2014 at 2:11 PM, Bram Moolenaar <[email protected]> wrote:
>
> Ben Fritz wrote:
>
>>
>> Finally, setting references in the LUA interface doesn't currently
>> allow aborting for failure using this patch. I could not figure out,
>> how to get a return value from lua_call. Can someone familiar with the
>> LUA interface code please help with this?
>
Still an issue. I think maybe I need to try a new thread.
>
> Thanks for making this patch!
>
> Did you run the tests under valgrind? That's a very good way to check
> for any memory access problems and leaks. Instructions are in
> src/testdir/Makefile. There are a few false positives, compare to a Vim
> without your patch.
>
I ran the tests ("make test", that is) under valgrind. A *few* false
positives?! Almost every test had at least one "definite leak" in the
output file.
I'm not sure I did it right, I did uncomment this line in the makefile:
+LEAK_CFLAGS = -DEXITFREE
and also the valgrind line from the test makefile.
Then I did "make clean", "make", "make test"
I copied off the test directory from before the patch and compared to
after the patch. For most test cases the only differences are in PID,
total heap size, and some address shifting.
For Test 16, there are plenty of differences in the "possible leak"
stuff. Before the test valgrind reports:
possibly lost: 1,210,076 bytes in 7,894 blocks
After the patch it becomes:
possibly lost: 1,224,143 bytes in 8,193 blocks
I'm really hoping that is a bunch of false positives; nothing looked
especially interesting but obviously I didn't take the time to examine
each line in detail. Test 16 is for the "secure" option which should
have nothing to do with the garbage collection changes, and most of
the supposed problems seem to have to do with the GUI startup.
I also used the makefile as a template to run valgrind on Vim while
running the crashtest.vim script that this patch fixes (after editing
counts to let it run in valgrind in faster than 2 years). The result
is exactly the same as every test except for test 16, except the heap
max size is a whole lot bigger.
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
