Fix memory leaks #32
Fix memory leaks #32
Conversation
There was a problem hiding this comment.
These are interesting findings 👍 For at least half of this, I'm not sure yet whether this is already the complete picture and the fix needed. This grep result puzzles me:
# git rev-parse HEAD ; git --no-pager grep hashstr_buf_size
4d37fd3549a199786796bd9bffd907523f5aea4d
src/hash.c:size_t hashstr_buf_size;
src/hash.h:extern size_t hashstr_buf_size;
src/hash.h: size_t hashstr_buf_size;
src/verify.c: cmp = memcmp(ihash, vhash, hashl->hash->hashstr_buf_size);
src/verify.c: cmp = memcmp(ihash, vhash, hashl->hash->hashstr_buf_size);
src/verify.c: i_hashstr_buf = malloc(hashstr_buf_size);
src/verify.c: v_hashstr_buf = malloc(hashstr_buf_size);
src/verify.c: cmp = memcmp(ihash, vhash, ihashlist->hash->hashstr_buf_size);When is hashstr_buf_size content — both globally and the struct member — ever written to?
|
Edit: hastype_t.final, which usee hashstr_buf, isn't written anywhere, as well. |
|
Hmm. For this one, I will leave it for the time when I'm able to run it through valgrind and gdb (probably over the weekend) and see whether data is ever written on those locations. |
|
@davidpolverari unfixed compile warnings in the CI logs seem to confirm. This is for latest |
|
Merged on master. Closing manually, as I had to rebase before merging. Thanks for the help! |
No description provided.