On Tue, Jan 12, 2016 at 06:00:49PM -0800, Eric Wheeler wrote:

Hello Eric,

quoted

Have you tested the patch below in SLE12-* when bcache is backed by md
raid5/6?

FYI: I was compared the drivers/md/bcache/io.c functions in the various
branches here:
  https://github.com/openSUSE/kernel
I compared the presence of bch_generic_make_request() (which the patch
below removes).  It looks like the branch SLE12-SP2 has the patch, but
version before SLE12-SP2 and openSUSE-* do not (as they still have
bch_generic_make_request).

Since the patch does exist in SLE12-SP2, I'm guessing that is been tested,
though I am curious if it has been tested specifically when being backed
by md raid5/6 so that queue->limits->partial_stripes_expensive is nonzero.

If you have and it is stable, then I want to get it to Kent and Jens for 
upstream integration.

The SLES12-SP2 kernel branch is very fresh, created this week. So while
the patch was tested by Johannes before adding it, and by our per-commit
automated tests, it didn't go through fully qualified QA test cycle yet.
So I won't say it's proven stable just yet.

Good to know, thank you for that!  

I encourage you to include md-based raid5/6 backed bcache volumes in your 
testing if it is not already.  This particular patch may affect that use 
case.  Please keep us posted, I look forward to learning about the 
bcache-specific test cycle results.

-Eric

-- 
Vojtech Pavlik
Director SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html