[Swan-dev] pluto: get rid of fake_md from aggr_outI1_continue()

[Andrew Cagney]

nice!

On Wed, 16 Jun 2021 at 03:03, D. Hugh Redelmeier <hugh at vault.libreswan.fi>
wrote:

> New commits:
> commit 0e67cf45937b7cd8b03c7bee24a3871dbb7618d6
> Author: D. Hugh Redelmeier <hugh at mimosa.com>
> Date:   Wed Jun 16 02:47:01 2021 -0400
>
>     pluto: get rid of fake_md from aggr_outI1_continue()
>
>     Stimulus: Coverity flagged a possible null pointer dereference of
>     fake_md->smc
>
>     I hadn't heard of this crash actually happening.  The reason is that
>     the "result" parameter for complete_v1_state_transition would normally
>     be STF_IGNORE and the function would return early with no crash.  In
>     theory, STF_INTERNAL_ERROR was possible but probably never happened.
>
>     I added a pexpect(e == STF_IGNORE) to aggr_outI1_continue.  This
>     should be replaced by appropriate logging.  Unfortunately I no longer
>     know how to log.
>
>     aggr_outI1_continue's call of complete_v1_state_transition is now
>     hardwired to use STF_IGNORE
>
>     complete_v1_state_transition's passert(md != NULL) was moved to avoid
>     demanding an md when none is needed.  (This passert seems silly since
>     all hardware we run on will catch null pointer dereferences.)
>

The log output is more readable.


>     At this point, fake_md became useless.  And also md's fake_dne field.
>
>     I cannot guarantee that Coverity will see that the NULL dereference is
>     eliminated.
>
> _______________________________________________
> Swan-commit mailing list
> Swan-commit at lists.libreswan.org
> https://lists.libreswan.org/mailman/listinfo/swan-commit
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.libreswan.org/pipermail/swan-dev/attachments/20210616/2a2c30b3/attachment.html>