I did a cursory review of the patch and the assocated code files (test-
buffer.c, buffer.c, and test-crypt.c).  I can't spot the source of the
problem; basically the source buffer being passed to buffer_copy() is
NULL in the Panic case.

The test_buffer_truncate_bits() function in test-buffer.c (which is the
caller of the crashing buffer_copy function) uses a static array of test
data. On ppc64el, the FIPS patch's changes to the cryptographic tests in
test-crypto.c are likely corrupting the memory of this static test data
array, which evidently NULL's the r_buffer pointer before the
test_buffer_truncate_bits() test runs.

This is probably a complicated bug that'll really need to be debugged on
ppc64el hardware that's able to repro it.  Meanwhile, probably worth
forwarding the issue upstream in case they have some better insights.

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2121250

Title:
  Panic: file buffer.c: line 366 (buffer_copy): assertion failed:
  (src->r_buffer != NULL)

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/dovecot/+bug/2121250/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to