From 4c89c0127dcadaa0d337c23d028f3fea5e76317a Mon Sep 17 00:00:00 2001 From: John Baldwin Date: Fri, 4 Aug 2023 16:41:05 -0700 Subject: g_raid concat: Fail requests to read beyond the end of the volume Previously a debug kernel would trigger an assertion failure if an I/O request attempted to read off the end of a concat volume, but a non-debug kernel would use an invalid sub-disk to try to complete the request eventually resulting in some sort of fault in the kernel. Instead, turn the assertions into explicit checks that fail requests beyond the end of the volume with EIO. For requests which run over the end of the volume, return a short request. PR: 257838 Reported by: Robert Morris Reviewed by: emaste MFC after: 1 week Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D41222 --- sys/geom/raid/tr_concat.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) (limited to 'sys') diff --git a/sys/geom/raid/tr_concat.c b/sys/geom/raid/tr_concat.c index eac62b3171d2..632501f323cd 100644 --- a/sys/geom/raid/tr_concat.c +++ b/sys/geom/raid/tr_concat.c @@ -240,8 +240,10 @@ g_raid_tr_iostart_concat(struct g_raid_tr_object *tr, struct bio *bp) offset -= vol->v_subdisks[no].sd_size; no++; } - KASSERT(no < vol->v_disks_count, - ("Request starts after volume end (%ju)", bp->bio_offset)); + if (no >= vol->v_disks_count) { + g_raid_iodone(bp, EIO); + return; + } bioq_init(&queue); do { sd = &vol->v_subdisks[no]; @@ -267,10 +269,8 @@ g_raid_tr_iostart_concat(struct g_raid_tr_object *tr, struct bio *bp) addr += length; offset = 0; no++; - KASSERT(no < vol->v_disks_count || remain == 0, - ("Request ends after volume end (%ju, %ju)", - bp->bio_offset, bp->bio_length)); - } while (remain > 0); + } while (remain > 0 && no < vol->v_disks_count); + bp->bio_completed = bp->bio_length - remain; while ((cbp = bioq_takefirst(&queue)) != NULL) { sd = cbp->bio_caller1; cbp->bio_caller1 = NULL; @@ -308,8 +308,8 @@ g_raid_tr_kerneldump_concat(struct g_raid_tr_object *tr, void *virtual, offset -= vol->v_subdisks[no].sd_size; no++; } - KASSERT(no < vol->v_disks_count, - ("Request starts after volume end (%ju)", boffset)); + if (no >= vol->v_disks_count) + return (EIO); do { sd = &vol->v_subdisks[no]; length = MIN(sd->sd_size - offset, remain); @@ -321,10 +321,9 @@ g_raid_tr_kerneldump_concat(struct g_raid_tr_object *tr, void *virtual, addr += length; offset = 0; no++; - KASSERT(no < vol->v_disks_count || remain == 0, - ("Request ends after volume end (%ju, %zu)", - boffset, blength)); - } while (remain > 0); + } while (remain > 0 && no < vol->v_disks_count); + if (remain > 0) + return (EIO); return (0); } @@ -340,7 +339,6 @@ g_raid_tr_iodone_concat(struct g_raid_tr_object *tr, g_destroy_bio(bp); pbp->bio_inbed++; if (pbp->bio_children == pbp->bio_inbed) { - pbp->bio_completed = pbp->bio_length; g_raid_iodone(pbp, pbp->bio_error); } } -- cgit v1.2.3