Valgrind is reporting multiple memory leaks #122
Comments
Some of these leaks are due to OpenSSL/libcurl, not really too much to worry about and besides we cannot do anything about them really.
Good catch, I am unaware of this one. Will fix :-)
This one is fixed on trunk I believe. Please try again with the trunk and debug info turned on, this way you get better traces from valgrind. Nico |
Result when building from trunk: |
herrhotzenplotz
added a commit
that referenced
this issue
Nov 27, 2023
The add_subcommand_alias routine contained a use-after-free bug which was discovered by valgrind. I actually used the AddressSanitizer to track down this bug - it is much more useful. The actual bug is that old_sc saved a reference to the old array that was realloc-ed and then the function pointer from out the now free'ed array was copied over. Fix this by storing the function pointer into a temporary, then realloc-ing the array and then copy the temporary value into the new array entry. Github-Issue: #122 Reported-By: Mikael Fangel (https://github.com/MikaelFangel)
herrhotzenplotz
added a commit
that referenced
this issue
Nov 27, 2023
The add_subcommand_alias routine contained a use-after-free bug which was discovered by valgrind. I actually used the AddressSanitizer to track down this bug - it is much more useful. The actual bug is that old_sc saved a reference to the old array that was realloc-ed and then the function pointer from out the now free'ed array was copied over. Fix this by storing the function pointer into a temporary, then realloc-ing the array and then copy the temporary value into the new array entry. Github-Issue: #122 Reported-By: Mikael Fangel (https://github.com/MikaelFangel)
herrhotzenplotz
added a commit
that referenced
this issue
Nov 27, 2023
The add_subcommand_alias routine contained a use-after-free bug which was discovered by valgrind. I actually used the AddressSanitizer to track down this bug - it is much more useful. The actual bug is that old_sc saved a reference to the old array that was realloc-ed and then the function pointer from out the now free'ed array was copied over. Fix this by storing the function pointer into a temporary, then realloc-ing the array and then copy the temporary value into the new array entry. Github-Issue: #122 Reported-By: Mikael Fangel (https://github.com/MikaelFangel)
herrhotzenplotz
added a commit
that referenced
this issue
Nov 28, 2023
I can't really derive all the leaks from the provided stack traces because they do not contain any debug information. Probably stripped or compiled without -g. Also, I can't reproduce with the same version of valgrind (3.22). Github-Issue: #122 Reported-By: Mikael Fangel (https://github.com/MikaelFangel)
|
I can't really derive any information from the stack trace because it contains a lot of noise and also it's missing debug info. Are the binaries stripped or are they not even compiled with -g and -O0? I have tried tracing down the leaks from the functions being called but I am unsure as I can't reproduce the leaks with the same version of valgrind locally (and another machine as well which ran the version 3.21). See the Can you please compile without all the Nix Shell noise and include the debug info (see HACKING.md). |
|
Fair, enough and yes, I actually forgot to compile with the debugging flag on. I'll see if I can set up a debugging environment during the week and provide some better info. (: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I've run a memory check with
valgrindon the application, and it looks like the application has several memory leaks.Simple example
Larger example
The text was updated successfully, but these errors were encountered: