svn commit: r326364 - head/sys/kern

[Hartmann, O.]

On Wed, 29 Nov 2017 12:49:22 +0000 (UTC)
Alexey Dokuchaev <danfe at FreeBSD.org> wrote:

> Author: danfe (ports committer)
> Date: Wed Nov 29 12:49:22 2017
> New Revision: 326364
> URL: https://svnweb.freebsd.org/changeset/base/326364
> 
> Log:
>   Fix several noticed style issues.
>   
>   Reviewed by:	bde
>   Approved by:	bapt
> 
> Modified:
>   head/sys/kern/subr_kdb.c
> 
> Modified: head/sys/kern/subr_kdb.c
> ==============================================================================
> --- head/sys/kern/subr_kdb.c	Wed Nov 29 10:04:11 2017
> (r326363) +++ head/sys/kern/subr_kdb.c	Wed Nov 29 12:49:22
> 2017	(r326364) @@ -250,7 +250,6 @@ kdb_reboot(void)
>   * its arguments.  Its up to the caller to ensure that the state
> variable is
>   * consistent.
>   */
> -
>  #define	KEY_CR		13	/* CR '\r' */
>  #define	KEY_TILDE	126	/* ~ */
>  #define	KEY_CRTLB	2	/* ^B */
> @@ -362,7 +361,6 @@ kdb_alt_break_gdb(int key, int *state)
>   * is selected or the current debugger does not support backtraces,
> this
>   * function silently returns.
>   */
> -
>  void
>  kdb_backtrace(void)
>  {
> @@ -410,7 +408,6 @@ kdb_backtrace_thread(struct thread *td)
>  /*
>   * Set/change the current backend.
>   */
> -
>  int
>  kdb_dbbe_select(const char *name)
>  {
> @@ -450,7 +447,6 @@ kdb_enter(const char *why, const char *msg)
>  /*
>   * Initialize the kernel debugger interface.
>   */
> -
>  void
>  kdb_init(void)
>  {
> @@ -485,7 +481,6 @@ kdb_init(void)
>  /*
>   * Handle contexts.
>   */
> -
>  void *
>  kdb_jmpbuf(jmp_buf new)
>  {
> @@ -521,9 +516,8 @@ kdb_reenter_silent(void)
>  }
>  
>  /*
> - * Thread related support functions.
> + * Thread-related support functions.
>   */
> -
>  struct pcb *
>  kdb_thr_ctx(struct thread *thr)
>  {
> @@ -617,7 +611,6 @@ kdb_thr_select(struct thread *thr)
>  /*
>   * Enter the debugger due to a trap.
>   */
> -
>  int
>  kdb_trap(int type, int code, struct trapframe *tf)
>  {
> _______________________________________________
> svn-src-head at freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/svn-src-head
> To unsubscribe, send any mail to
> "svn-src-head-unsubscribe at freebsd.org"

r326363 produces a panic while performing installworld. This happens
even with the recommended procedure in singleuser mode with booted
kerenl r326363. While installing manpages, the system went down
reporting a "spinlock held too long".

As a great result of that, booting the box results in a fancy screen
with register dumps from BTX loader and BTX halted. This leaves me now
with the second box completely crashed within two days with the same
serious bug! Well done.

As an emergency recovery I tried to replace /bin, /sbin, /libexec /lib
with binaries from another system of the same architecture (Ivybridge
XEON), from 27th of November (it is a system compiled around 21 h local
German time, I have no access to those systems anymore as I try to
explain). In one case this rescue did not work - neither the old nor
the new kernel boot, they get stuck. 

Since all systems do have a "clean from scratch build" of CURRENT
world, this infrastructure is not accessible anymore sinde the
rediculous reduced images of FreebSD is useless for system recovery via
installworld/installkernel since any necessary compiler infrastructure
is now erased.

What to do? How can I fix this? I need at least one box working again
to repair the others.

Thanks in advance and thank you very much.

oh