[Bug 263853] geom(4): raid/md_jmicron.c: jmicron RAID taste code can panic if conf is garbage

2024-10-04 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=263853

Mark Linimon  changed:

   What|Removed |Added

   Keywords||patch
  Flags|mfc-stable13?,  |
   |mfc-stable12?   |

--- Comment #6 from Mark Linimon  ---
^Triage: clear unneeded flags.  Nothing has yet been committed to be merged.

While here, set 'patch' flag to indicate inline patch (which does not seem to
have been applied.  Let us know if that is not correct).

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 238814] geom: topology lock being dropped in dumpconf of gate, raid, & raid3

2024-10-04 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=238814

--- Comment #6 from Mark Johnston  ---
(In reply to Mark Johnston from comment #5)
I spent a bit of time looking at this, but for g_raid it's quite non-trivial
and hard to test.  My patch for gmirror also relaxes synchronization perhaps a
bit too much, in that inconsistencies are possible (e.g., disk state
transitions and syncid/genid bumps happen without locking), though it fixes the
larger problem.

(In reply to Mark Johnston from comment #5)
My comment about GEOM gate is wrong, the topology lock simply can't be dropped.
 Fortunately, that one is easy to fix.

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 238814] geom: topology lock being dropped in dumpconf of gate, raid, & raid3

2024-10-04 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=238814

--- Comment #7 from commit-h...@freebsd.org ---
A commit in branch main references this bug:

URL:
https://cgit.FreeBSD.org/src/commit/?id=b37b2543a23b44c78f6d78823dcfcedba46570db

commit b37b2543a23b44c78f6d78823dcfcedba46570db
Author: Mark Johnston 
AuthorDate: 2024-10-04 14:53:57 +
Commit: Mark Johnston 
CommitDate: 2024-10-04 15:56:34 +

ggate: Avoid dropping the GEOM topology lock in dumpconf

In general it's not safe to drop the topology lock in these routines, as
GEOM assumes that the mesh will be consistent during traversal.
However, there's no reason we can't hold the topology lock across calls
to g_gate_release().  (Note that g_gate_hold() can be called with the
topology lock held.)

PR: 238814
MFC after:  2 weeks

 sys/geom/gate/g_gate.c | 3 ---
 1 file changed, 3 deletions(-)

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 238814] geom: topology lock being dropped in dumpconf of gate, raid, & raid3

2024-10-04 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=238814

--- Comment #5 from Mark Johnston  ---
(In reply to Mark Linimon from comment #4)
Yes, the problem still exists.  I think the gate GEOM is okay, but the patch
from comment 1 needs to be adapted to the raid and raid3 GEOMs.

-- 
You are receiving this mail because:
You are the assignee for the bug.