* [PATCH] cont_prepare_write question
@ 2003-07-21 19:45 Eli Carter
0 siblings, 0 replies; only message in thread
From: Eli Carter @ 2003-07-21 19:45 UTC (permalink / raw)
To: linux-fsdevel
From a very basic look at cont_prepare_write() in fs/buffer.c, it looks
like it needs this patch to prevent an infinite loop when extending a
file by several blocks. (>8 for the 0.5k block case)
--- linux-2.5.68-rmk1/fs/buffer.c 2003/04/22 15:13:04 1.1.2.3
+++ linux-2.5.68-rmk1/fs/buffer.c 2003/07/21 19:38:40
@@ -2118,6 +2118,7 @@
goto out_unmap;
kaddr = kmap_atomic(new_page, KM_USER0);
memset(kaddr+zerofrom, 0, PAGE_CACHE_SIZE-zerofrom);
+ *bytes += PAGE_CACHE_SIZE-zerofrom;
flush_dcache_page(new_page);
kunmap_atomic(kaddr, KM_USER0);
__block_commit_write(inode, new_page,
I'm not going to pretend that I understand why, or what this exactly
accomplishes... it just seems to improve things somewhat in my case. If
someone could give a basic explanation of what is going on in this loop,
particularly in regards to how the loop is supposed to end, or a pointer
to something to read that would make this function clearer than mud, I'd
be very grateful.
Thanks,
Eli
--------------------. "If it ain't broke now,
Eli Carter \ it will be soon." -- crypto-gram
eli.carter(a)inet.com `-------------------------------------------------
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2003-07-21 19:30 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-07-21 19:45 [PATCH] cont_prepare_write question Eli Carter
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox