[ale] Calling all Linux users!

[Todd Graham Lewis]

2.0, while a great step beyond 1.2, seems to have been plagued by
stability problems and overall bad luck.  David Miller just flat gave up
on getting a good 2.0.31 out the door, and Linus stepped up to the task.
It has not been smooth going.

Werner has done a lot of neat work optimizing the buffer cache, and
according to credible sources within a certain large, nay national, ISP,
his patches completely rock.  

If you've ever compiled a kernel before, then please look into trying
out Linus' pre-2.0.31 patches; if you haven't, then read the docs or
ask here and figure out how.  Try to break them.  Try them with wierd
hardware and shoot your machine's load up to 40.  Remember, Microsoft
just has to sneeze and they get a gazillion beta testers begging for M$
code to screw up their week, and if Linux is about anything, it's about
keeping up with the Joneses.

If anyone thinks they've discovered a bug in pre-2.0.31, then please
feel free to report it here before taking it to linux-kernel.  There are
enough kernel-knowledgeable people on this list to weed out misinformed
bug reports, and once those are taken care of then the rest are a huge
help to the kernel folk.

Go ALE!

-- 
Todd Graham Lewis       Manager of Web Engineering    MindSpring Enterprises
(800) 719-4664, x2804             Linux!               tlewis at mindspring.net

---------- Forwarded message ----------
Date: Mon, 18 Aug 1997 21:06:47 +0200
 From: "Dr. Werner Fink" <werner at suse.de>
To: gadio at netvision.net.il
Cc: torvalds at transmeta.com, whawes at star.net, mantel at suse.de,
    linux-kernel at vger.rutgers.edu
Newsgroups: internal.lists.linux.kernel
Subject: Smooth buffer for 2.0.31 [Was: kflushd() patch]


> The following patch fixes a busy-loop deadlock in kflushd(). This
> deadlock was probably only triggered under rare conditions (if ever),
> but as a side effect of my recent buffer.c patch, it is more common.
> 
> The deadlock can occur if kflushd() finds all the buffers (most
> likely just one :-) on the dirty list to be both dirty and locked:

[...]

I've added a small change to your patch. It simply allows growing
buffer with GFP_BUFFER but only if absolute necessary. I've found this
solution experimentally by playing with some conditions and positions
calling wakeup_bdflush(1) in refill_freelist. This is done by running
two `bonnie -s 200' and two `make clean; make -j zImage' in loops.

The patch is against pre-patch-2.0.31-7 plus Gadi's deadlock patch ...
it's one of the fastest kernel I've ever seen.  ... But there are
also testers needed for this patch _before_ the next pre-patch release
or the real 2.0.31 ... anybody out there?


           Werner

-----------------------------------------------------------------------------
diff -urN linux-2.0.31-linus/fs/buffer.c linux/fs/buffer.c
--- linux-2.0.31-linus/fs/buffer.c	Mon Aug 18 13:58:51 1997
+++ linux/fs/buffer.c	Mon Aug 18 18:43:32 1997
@@ -662,6 +662,12 @@
 		}
 		goto repeat;
 	}
+
+	/* Dirty buffers should not overtake, wakeup_bdflush(1) calls
+	   bdflush and sleeps, therefore kswapd does his important work. */
+	if ((nr_buffers_type[BUF_DIRTY] > nr_buffers * bdf_prm.b_un.nfract/100) ||
+	    (nr_free_pages < (min_free_pages >> 1)))
+		wakeup_bdflush(1);
 	
 	/* Too bad, that was not enough. Try a little harder to grow some. */
 	
@@ -672,19 +678,8 @@
 		};
 	}
 
-#if 0
-	/*
-	 * In order to protect our reserved pages, 
-	 * return now if we got any buffers.
-	 */
-	if (free_list[BUFSIZE_INDEX(size)])
-		return;
-
 	/* and repeat until we find something good */
-	if (!grow_buffers(GFP_ATOMIC, size))
-		wakeup_bdflush(1);
-#endif
-	wakeup_bdflush(1);
+	grow_buffers(GFP_ATOMIC, size);
 
 	/* decrease needed even if there is no success */
 	needed -= PAGE_SIZE;
@@ -1717,7 +1712,7 @@
 		 * dirty buffers, then make the next write to a
 		 * loop device to be a blocking write.
 		 * This lets us block--which we _must_ do! */
-		if (ndirty == 0 && nr_buffers_type[BUF_DIRTY] > 0) {
+		if (ndirty == 0 && nr_buffers_type[BUF_DIRTY] > 0 && wrta_cmd != WRITE) {
 			wrta_cmd = WRITE;
 			continue;
 		}
@@ -1725,7 +1720,7 @@
 		
 		/* If there are still a lot of dirty buffers around, skip the sleep
 		   and flush some more */
-		if(nr_buffers_type[BUF_DIRTY] <= nr_buffers * bdf_prm.b_un.nfract/100) {
+		if(ndirty == 0 || nr_buffers_type[BUF_DIRTY] <= nr_buffers * bdf_prm.b_un.nfract/100) {
 			wake_up(&bdflush_done);
 			current->signal = 0;
 			interruptible_sleep_on(&bdflush_wait);