The pt_magic field is not the first one in the pthread_t structure.
After this fix, this code is confirmed to work and function td_thr_info()
is functional.
Sponsored by <The NetBSD Foundation>
The following PR are now resolved and are reported to be properly fixed:
PR lib/51635 td_thr_iter in <pthread_dbg.h> seems broken
PR lib/51633 tests/lib/libpthread_dbg/t_dummy unreliable
Sponsored by <The NetBSD Foundation>
i386.
The PTE space is a critical region that maps the page tree, and bugs have
been found in both amd64 and i386 where the kernel would wrongly overflow
userland data on this area. This kind of bug is terrible, since it allows
userland to overwrite some entries of the page tree, which makes it easy
to patch the kernel text and get ring0 privileges.
parameters/enables for error recovery, similar to {get,set}cache.
Many old SCSI disks shipped with reallocation disabled, albeit supporting it.
Minor (cosmetic) fixup of scsi_disk_pages while there.
Based upon code in PR bin/29165 by Greg A. Woods.
OK christos@
Source and destination were swapped. The source of this confusion was that
running these tests under gdb(1) will generate false positives as it will
initialize pthread__dbg to PID of the debugger. This means that it is
currently not possible to debug pthread_dbg code under a full-stack
debugger using NetBSD debugging library for threads.
This should address:
PR lib/51633 tests/lib/libpthread_dbg/t_dummy unreliable
PR lib/51635: td_thr_iter in <pthread_dbg.h> seems broken
After applying the fix I'm able to run all pthread_dbg tests without
indeterminism. The indeterminism was caused by overwritting source of data
with trash.
Sponsored by <The NetBSD Foundation>
(reported by rhialto@falu.nl) Don't fail to start if it doesn't.
Make sure the directory for the config file exists inside the chroot before
attempting to copy into it ("confdir" was calculated, but never used...)
While here, fix getopts usage (obviously only ever previously tested when
the -c arg was the first option...) and don't use test(1)'s -o operator
(especially not when one of the other args is an unknown string).
If -c is given (and we will chroot), require filename to be full path.
Misc minor style cleanups.
This test contains threads1 test that:
Asserts that td_thr_iter() call without extra logic works
This tests fails and is linked with gnats:
PR lib/51635: td_thr_iter in <pthread_dbg.h> seems broken
Sponsored by <The NetBSD Foundation>
to preserve bits like KX in case we catch an interrupt between mtc0 and the
write actually taking effect
now n32 kernels on my O2 are (mostly) stable again
ok skrll@
And then, if we want to re-insert the removed pslist element, we need to
call PSLIST_ENTERY_INIT again.
advised by riastradh@n.o and reviewed by ozaki-r@n.o, thanks.
This problem occurs only if NET_MPSAFE on.
ifconfig destroy side:
kernel entry point is ifioctl => if_clone_destroy.
pr_purgeif() acquires softnet_lock, and then ifa_remove() calls
pserialize_perform() holding softnet_lock.
ifconfig side:
kernel entry point is socreate.
pr_attach()(udp_attach_wrapper()) calls sosetlock(). In this call path,
sosetlock() try to acquire softnet_lock.
These can cause dead lock.
The dummy2 test reports failures on i386 releng machines.
The same with dummy3, but it's also reproducible sometimes on my amd64
setup. It's not always broken, neither always functional - once a while it
reports failure. So far I have not reproduced failure in dummy2.
PR lib/51633 tests/lib/libpthread_dbg/t_dummy unreliable
Sponsored by <The NetBSD Foundation>
