[PATCH] Fix hanging close for /dev/mtd

[Jörn Engel]

[ Putting David on Cc:.  This is a bug report for JFFS2 ]

On Wed, 13 June 2007 15:29:38 +0200, Joakim Tjernlund wrote:
> On Wed, 2007-06-13 at 13:54 +0200, Jörn Engel wrote:
> > 
> > In this case you made me wonder.  What is the mtd->sync method supposed
> > to accomplish anyway?  mtd->write is synchronous, mtd->erase de-facto
> > also is.  So it cannot be a sync in the sense of "flush all dirty data
> > to flash and don't return before you're finished".  Or can it?
> 
> Yes, this is also my view(for NOR flash, NAND is diffrent).
> mtd->erase can be suspended though, but it doesn't matter
> in this case. However, I don't know enough to be 100% sure

At the very least that method needs better documentation than:
	/* Sync */
	void (*sync) (struct mtd_info *mtd);

I can see two sane alternatives.

/*
 * An MTD driver may return from mtd->write() calls before the data has
 * safely been written to the medium.  Any amount of dirty data may be
 * cached by either the driver or the device.  If a filesystem depends
 * on the data being safely on the medium, it has to explicitly call
 * mtd->sync().
 */

/*
 * Mtd->write() and mtd->erase() methods may only return if the data has
 * been safely written to the device.  Special care needs to be taken
 * for devices that support caching.
 */

My tendency is to favor the first variant.  Block2mtd already implements
it.  And if we went for variant two, mtd->sync() would be pointless and
should get removed.

Now to the bug report:
JFFS2 must call mtd->sync() for its various sync routines.  If mounted
with -o sync or for files opened with O_SYNC, it has to call that after
every write.  If nothing else, it is currently broken on block2mtd.

Jörn

-- 
Linux [...] existed just for discussion between people who wanted
to show off how geeky they were.
-- Rob Enderle