https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=297512
--- Comment #3 from Stéphane D'Alu <[email protected]> --- Tested D58838 on my machine, the screen unlock problem didn't reproduced. Thanks for the quick fix! Here is the follow-up of the test suite run by Fable: - Whole-system scan (ptrace-based measurement of each process's last-env-string distance to the next unmapped page): 96 processes, 3 landed in the previously-failing layout (distance < 256 with the next page unmapped: zfsd at 163, getty at 164, ibus-portal at 203). kern.proc.env now succeeds for all of them; zero ENOMEM system-wide. On the unpatched kernel the same predicate matched the ENOMEM failures exactly (4 predicted / 4 failing out of 139). - On-demand reproduction: spawned /bin/sleep with varying environment sizes until an instance landed 72 bytes from the unmapped boundary; procstat penv read its complete environment (33 entries) without error. -- You are receiving this mail because: You are the assignee for the bug.
