[PATCH] jffs2: Fix lock acquisition order bug in jffs2_write_begin

[Thomas.Betker at rohde-schwarz.com]

Hello Joakim,

I am not sure if it is okay to call jffs2_complete_reservation() without a 
previous jffs2_reserve_space(), so I have avoided that in my own patch.

Also, I am not happy with using ret to check if 
jffs2_complete_reservation() was done; I am using alloc_len for that.

"jffs2_dbg(1, "%s()\n", __func__);" comes a bit late.

And "be be" in the first comment should probably read "be done".

I am attaching my own patch for comparison (sorry, I haven't found yet a 
way to send the patch inline); the tests are running this morning.

Best regards,
Thomas

-------------- next part --------------
A non-text attachment was scrubbed...
Name: jffs2_write_begin-v2.patch
Type: application/octet-stream
Size: 2569 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-mtd/attachments/20121008/2fa5b545/attachment.obj>