14 Nov, 2005
1 commit
-
Signed-off-by: Thomas Gleixner
08 Nov, 2005
1 commit
-
Some manual fixups for clashing kfree() cleanups etc.
07 Nov, 2005
34 commits
-
This is the fs/ part of the big kfree cleanup patch.
Remove pointless checks for NULL prior to calling kfree() in fs/.
Signed-off-by: Jesper Juhl
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Signed-off-by: Thomas Gleixner
-
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Some callers to block-layer commit_write function treat non-zero return as
error, notably the loopback mount driver sometimes used in conjunction with
JFFS2 on NAND flash for bad block avoidance, etc. Return zero for success
as do various other commit_write functions.Signed-off-by: Todd Poynor
Signed-off-by: Thomas Gleixner -
- assume wbuf may be of size which is not power of 2
- don't make strange assumption about not padding wbuf for DataFlash
- use wbuf = DataFlash page and eraseblock >= 8 Dataflash pagesFrom: Peter Menzebach
Acked-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Simplify the debugging code further.
Update the TODO listSigned-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Always keep valid data in reserved_size.
It did not cause problems, but the reservation code was unoptimal
when centralized summary was active or the size of the erase block
was very small.Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Do the summary collection in the right place. If the device
was not writebuffered but had c->mtd->writev function
(e.g. blkmtd) the summary collector function was not called.Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
The goal of summary is to speed up the mount time. Erase block summary (EBS)
stores summary information at the end of every (closed) erase block. It is
no longer necessary to scan all nodes separetly (and read all pages of them)
just read this "small" summary, where every information is stored which is
needed at mount time.This summary information is stored in a JFFS2_FEATURE_RWCOMPAT_DELETE. During
the mount process if there is no summary info the orignal scan process will
be executed. EBS works with NAND and NOR flashes, too.There is a user space tool called sumtool to generate this summary
information for a JFFS2 image.Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Remove support for virtual blocks, which are build by
concatenation of multiple physical erase blocks.For more information please read the MTD mailing list thread
"[PATCH] remove support for virtual blocks"Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
When data starts from the beginning of NAND page, 'len' must be zero, not
c->wbuf_page.Thanks to Zoltan Sogor for reporting this problem.
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
From: David Woodhouse
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Intels Sibley flash needs JFFS2 write buffer functionality
Signed-off-by: Nicolas Pitre
Signed-off-by: Thomas Gleixner -
Remove more noisy debugs. Add current->pid to debug messages.
Remove bogus includes.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
The previous changes introduced wrong length calculations.
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Remove extra noisy debugs
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Instead of building fragtree starting from node with the smallest version
number, start from the highest. This helps to avoid reading and checking
obsolete nodes.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
The debug code cleanup broke the eCos build.
Signed-off-by: Andrew Lunn
Signed-off-by: Thomas Gleixner -
Replace the D1(printk()) style debugging with the new debug macros
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Move functions to read inodes into readinode.c
Move functions to handle fragtree and dentry lists into nodelist.[ch]Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Small comment cleanups. Remove a unused macro
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Remove unused compressor code
Signed-off-by: Ferenc Havasi
Signed-off-by: Thomas Gleixner -
Rename functions to a name matching the functionality.
Remove stall debug codeSigned-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Various simplifiactions. printk format corrections.
Convert more code to use the new debug functions.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner
06 Nov, 2005
4 commits
-
When JFFS22 is unable to read the root inode, the bad root inode object is not
freed and remains sticked in the jffs2_i slab cache. When we further try to
free the slab cache (e.g., on rmmod jffs2), slab allocator subsystem panics.
Fix this bug.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
If debugging is disabled, define debugging functions as empty macros, instead
of using Dx() explicitly.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
JFFS2 uses f->dents to store the pointer to the symlink target string (in case
the inode is symlink). This is somewhat ugly to use the same field for
different reasons. Introduce distinct field f->target for this purpose.
Note, f->fragtree, f->dents, f->target may probably be put in a union.Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner -
Move debug functions into a seperate source file
Signed-off-by: Artem B. Bityutskiy
Signed-off-by: Thomas Gleixner