Date: Mon, 13 Jul 2015 04:33:16 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-bugs@FreeBSD.org Subject: [Bug 201517] Kernel hang on boot with KDTRACE_HOOKS enabled after r284998 Message-ID: <bug-201517-8@https.bugs.freebsd.org/bugzilla/>
next in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201517 Bug ID: 201517 Summary: Kernel hang on boot with KDTRACE_HOOKS enabled after r284998 Product: Base System Version: 10.1-STABLE Hardware: Any OS: Any Status: New Severity: Affects Some People Priority: --- Component: kern Assignee: freebsd-bugs@FreeBSD.org Reporter: matthew@reztek.cz Updating world on a 10-STABLE machine to the point of 10.2-BETA in order to catch the last OpenSSL update resulted in a kernel that hangs on boot. The hang occurs approximately when USB buses should be probed. The floppy drive controller was probed and the drive is still spinning (stuck with light on), firewire ports are detected, and sometimes but not always the USB ports are detected, but never gets as far as detecting other disks. Kernel must be booted without any drm2 drivers loaded in order to see where it hangs. If KMS init has started the display outputs are still off (no signal, monitor in powersave) when the kernel hangs. Bisecting (manually) the commits between the versions and test booting a dozen or so kernels revealed the lockstat improvements MFHed from r284297 in r284998 are the culprit. The version before that commit is fine, but that exact revision and any after hang on boot as described. Current revision with that changeset reversed works ok. Current revision with all changesets works ok if KDTRACE_HOOKS is disabled, so the offending change must be in one of the sections that is #ifdef KDTRACE_HOOKS. -- You are receiving this mail because: You are the assignee for the bug.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-201517-8>