summaryrefslogtreecommitdiff
path: root/ubi-utils/TODO
diff options
context:
space:
mode:
authorFrank Haverkamp <haver@vnet.ibm.com>2008-01-07 14:05:00 +0100
committerArtem Bityutskiy <Artem.Bityutskiy@nokia.com>2008-01-15 12:47:00 +0200
commit4bf037f3a2aea4328e10422379f751370d288d46 (patch)
tree68a24f007e8b89609d14d44165a8bcd669629e7a /ubi-utils/TODO
parent07e6203fd3078d6855e7d33597bdbbae10e555fd (diff)
ubi-utils: bin2nand, nand2bin add support of different ecc layouts
Both tools were lacking support of alternate ECC layouts. Only our intitial format was supported. With this change, it should be very easy to add more layouts in addtion to the already supported ones, which are: IBM (our format), and the MTD default layout. NAND OOB sizes of 512 and 2048 are currently supported. In contrast to the old version of bin2nand, the holes inbetween the ECC data is now not filled with 0x00 anymore but instead 0xff (like deleted flash) is used. This should not cause any difference. The testcase reflects the different layouts too. Signed-off-by: Frank Haverkamp <haver@vnet.ibm.com>
Diffstat (limited to 'ubi-utils/TODO')
-rw-r--r--ubi-utils/TODO7
1 files changed, 5 insertions, 2 deletions
diff --git a/ubi-utils/TODO b/ubi-utils/TODO
index a604170..4683bd7 100644
--- a/ubi-utils/TODO
+++ b/ubi-utils/TODO
@@ -3,12 +3,15 @@ TODO
* Range checking is broken, reserving 2M and offering 3M binary data
... works!? No!
- * Remove the above misterious statement or make it understandable
+ * Remove the above mysterious statement or make it understandable
* doc/ directory contains a file which refers images from /home/frank/
- please fix this or remove the whole doc/ altogether.
* the tests from the scripts/ directory should live in
- mit-utils/tests/ubi-tests/ and it would be nice to have a short
+ mtd-utils/tests/ubi-tests/ and it would be nice to have a short
description of the tests
* the stuff from the perl/ directory should go to the scripts/
* May we pleas split UBI-related stuff and pure NAND-related stuff and
not to keep this all in one.
+
+ * nand2bin and bin2nand need more flexibility to influence the ECC
+ data placement.