| commit | d9ba2e1a75f7addc184949801b561b026dd26613 | |
| author | Dan Carpenter <error27@gmail.com> | |
| Wed, 8 Mar 2023 11:25:36 +0000 (8 14:25 +0300) | ||
| committer | Dan Carpenter <error27@gmail.com> | |
| Wed, 8 Mar 2023 11:25:36 +0000 (8 14:25 +0300) | ||
| tree | 73cbf24305ee327bdb463b22e5c8f4fa538af270 | treesnapshot (tar.gz zip) |
| parent | e586ba093e8a13b5bcd2f0149be8fa06d736af41 | commitdiff |
param_limit: ignore BUF_CLEAR here as well
This change is slightly less clear, but a singly PARAM_CLEAR of a massive
struct can end up creating over 1000 param_limit states so it can cause
quite a bit of slow down.
The param_limit states are used for understanding which return paths are
possible. So I imagine this might cause some false positives. But I doubt
it causes enough false positives to notice.
Signed-off-by: Dan Carpenter <error27@gmail.com>
This change is slightly less clear, but a singly PARAM_CLEAR of a massive
struct can end up creating over 1000 param_limit states so it can cause
quite a bit of slow down.
The param_limit states are used for understanding which return paths are
possible. So I imagine this might cause some false positives. But I doubt
it causes enough false positives to notice.
Signed-off-by: Dan Carpenter <error27@gmail.com>
| smatch_param_limit.c | diffblobblamehistory |