[01/15] UBI: Fastmap: Relax EBA selfcheck

Submitted by Richard Weinberger on June 21, 2012, 9:49 p.m.


Message ID 1340315408-27375-2-git-send-email-richard@nod.at
State New, archived
Headers show

Commit Message

Richard Weinberger June 21, 2012, 9:49 p.m.
It is allowed that fastmap knows an EBA table entry
as unmapped while scanning would know this entry as
Fastmap knows the current EBA table and scannning
reconstructs it. This case can happen if the fastmap
gets written while the entry is alredy unmapped but the
PEB has not been erased.
If the UBI device gets detached at this moment, scanning
may detect this EBA entry as mapped.
(Of course the upper layer, UBIFS, knows this LEB is no longer

Signed-off-by: Richard Weinberger <richard@nod.at>
 drivers/mtd/ubi/eba.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

Patch hide | download patch | download mbox

diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c
index 2bee8be..6698d5e 100644
--- a/drivers/mtd/ubi/eba.c
+++ b/drivers/mtd/ubi/eba.c
@@ -1276,7 +1276,8 @@  int self_check_eba(struct ubi_device *ubi, struct ubi_attach_info *ai_fastmap,
 		for (j = 0; j < vol->reserved_pebs; j++) {
 			if (scan_eba[i][j] != fm_eba[i][j]) {
-				if (scan_eba[i][j] == UBI_LEB_UNMAPPED)
+				if (scan_eba[i][j] == UBI_LEB_UNMAPPED ||
+					fm_eba[i][j] == UBI_LEB_UNMAPPED)
 				ubi_err("LEB:%i is mapped to PEB:%i instead of PEB:%i!", i, fm_eba[i][j], scan_eba[i][j]);