Re: [PATCH V2] raid1: ensure write behind bio has less than BIO_MAX_VECS sectors

From
Guoqing Jiang <hidden>
Date
2021-08-20 08:19:33
Also in
linux-raid
Reply to
Christoph Hellwig, 2021-08-19

On 8/19/21 4:55 PM, Christoph Hellwig wrote:
On Wed, Aug 18, 2021 at 03:37:38PM +0800, Guoqing Jiang wrote:
quoted
  	for (i = 0;  i < disks; i++) {
  		struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
+
+		if (test_bit(WriteMostly, &mirror->rdev->flags))
+			write_behind = true;
How does this condition relate to the ones used for actually calling
alloc_behind_master_bio?  It looks related, but as someone not familiar
with the code I can't really verify if this is correct, so a comment
explaining it might be useful.
How about this?

+               /*
+                * The write-behind io is only attempted on drives marked as
+                * write-mostly, which means we will allocate write behind
+                * bio later.
+                */
                if (test_bit(WriteMostly, &mirror->rdev->flags))
                        write_behind = true;
quoted
+	/*
+	 * When using a bitmap, we may call alloc_behind_master_bio below.
+	 * alloc_behind_master_bio allocates a copy of the data payload a page
+	 * at a time and thus needs a new bio that can fit the whole payload
+	 * this bio in page sized chunks.
+	 */
+	if (write_behind && bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
Overly long line here.
I can change it given you still prefer the  limitation is 80 characters.

Thanks,
Guoqing
Thread (7 messages) 7 messages, 3 authors, last reply 2021-08-23
Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help