| commit | 4df485caecfb5dd984e2e29084ef959297aef45f | |
| author | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 12 Mar 2018 12:42:33 +0000 (12 15:42 +0300) | ||
| committer | Dan Carpenter <dan.carpenter@oracle.com> | |
| Mon, 12 Mar 2018 12:42:33 +0000 (12 15:42 +0300) | ||
| tree | 9444741fcd5a62668339408447b8e352ba5d43c2 | treesnapshot (tar.gz zip) |
| parent | 334907b60d00238c5e3482ffbf87b1889ffa9ce8 | commitdiff |
ranges/ptrlist: solve a leak with a big hack (second attempt)
The truth is that I don't know why the record_rl() way of freeing all the
rl ptr lists wasn't working. I really thought it should have worked.
Anyway, the new "fix" is to create a special ptr_list allocator for rl
pointer lists and then use the clear_ptrlist_alloc() to free it. It seems
to solve the issue I was having with running out of memory running
smatch --info on the kernel.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
The truth is that I don't know why the record_rl() way of freeing all the
rl ptr lists wasn't working. I really thought it should have worked.
Anyway, the new "fix" is to create a special ptr_list allocator for rl
pointer lists and then use the clear_ptrlist_alloc() to free it. It seems
to solve the issue I was having with running out of memory running
smatch --info on the kernel.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
| ptrlist.c | diffblobblamehistory | |
| smatch_ranges.c | diffblobblamehistory |