Re: [PATCH 2/2] zram: user per-cpu compression streams

[Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>]

On (05/03/16 11:30), Sergey Senozhatsky wrote:
> > We are concerning about returing back to no per-cpu options but actually,
> > I don't want. If duplicate compression is really problem(But It's really
> > unlikely), we should try to solve the problem itself with different way
> > rather than roll-back to old, first of all.
> > 
> > I hope we can. So let's not add big worry about adding new dup stat. :)
> 
> ok, no prob. do you want it a separate sysfs node or a column in mm_stat?
> I'd prefer mm_stat column, or somewhere in those cumulative files; not a
> dedicated node: we decided to get rid of them some time ago.
> 

will io_stat node work for you?
I'll submit a formal patch later today. when you have time, can you
take a look at http://marc.info/?l=linux-kernel&m=146217628030970 ?
I think we can fold this one into 0002. it will make 0002 slightly
bigger, but there nothing complicated in there, just cleanup.

====

From: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Subject: [PATCH] zram: export the number of re-compressions

Make the number of re-compressions visible via the io_stat node,
so we will be able to track down any issues caused by per-cpu
compression streams.

Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Suggested-by: Minchan Kim <minchan@kernel.org>
---
 Documentation/blockdev/zram.txt | 3 +++
 drivers/block/zram/zram_drv.c   | 7 +++++--
 drivers/block/zram/zram_drv.h   | 1 +
 3 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/Documentation/blockdev/zram.txt b/Documentation/blockdev/zram.txt
index 5bda503..386d260 100644
--- a/Documentation/blockdev/zram.txt
+++ b/Documentation/blockdev/zram.txt
@@ -183,6 +183,8 @@ mem_limit         RW    the maximum amount of memory ZRAM can use to store
 pages_compacted   RO    the number of pages freed during compaction
                         (available only via zram<id>/mm_stat node)
 compact           WO    trigger memory compaction
+num_recompress    RO    the number of times fast compression paths failed
+                        and zram performed re-compression via a slow path
 
 WARNING
 =======
@@ -215,6 +217,7 @@ whitespace:
 	failed_writes
 	invalid_io
 	notify_free
+	num_recompress
 
 File /sys/block/zram<id>/mm_stat
 
diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
index 817e511..11b19c9 100644
--- a/drivers/block/zram/zram_drv.c
+++ b/drivers/block/zram/zram_drv.c
@@ -395,7 +395,8 @@ static ssize_t io_stat_show(struct device *dev,
 			(u64)atomic64_read(&zram->stats.failed_reads),
 			(u64)atomic64_read(&zram->stats.failed_writes),
 			(u64)atomic64_read(&zram->stats.invalid_io),
-			(u64)atomic64_read(&zram->stats.notify_free));
+			(u64)atomic64_read(&zram->stats.notify_free),
+			(u64)atomic64_read(&zram->stats.num_recompress));
 	up_read(&zram->init_lock);
 
 	return ret;
@@ -721,8 +722,10 @@ compress_again:
 
 		handle = zs_malloc(meta->mem_pool, clen,
 				GFP_NOIO | __GFP_HIGHMEM);
-		if (handle)
+		if (handle) {
+			atomic64_inc(&zram->stats.num_recompress);
 			goto compress_again;
+		}
 
 		pr_err("Error allocating memory for compressed page: %u, size=%zu\n",
 			index, clen);
diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h
index 06b1636..78d7e8f 100644
--- a/drivers/block/zram/zram_drv.h
+++ b/drivers/block/zram/zram_drv.h
@@ -85,6 +85,7 @@ struct zram_stats {
 	atomic64_t zero_pages;		/* no. of zero filled pages */
 	atomic64_t pages_stored;	/* no. of pages currently stored */
 	atomic_long_t max_used_pages;	/* no. of maximum pages stored */
+	atomic64_t num_recompress; /* no. of failed compression fast paths */
 };
 
 struct zram_meta {
-- 
2.8.2