From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <axboe@kernel.dk>
Subject: Re: [PATCH 16/17] mtd_blkdevs: convert to blk-mq
To: Richard Weinberger <richard.weinberger@gmail.com>
Cc: linux-block@vger.kernel.org,
 "linux-mtd @ lists . infradead . org" <linux-mtd@lists.infradead.org>,
 David Woodhouse <dwmw2@infradead.org>
References: <20181011165909.32615-1-axboe@kernel.dk>
 <20181011165909.32615-17-axboe@kernel.dk>
 <CAFLxGvy04s++HExKixU02=e86kx4yqcz+boeepxi9=uDU94sQg@mail.gmail.com>
From: Jens Axboe <axboe@kernel.dk>
Message-ID: <263ca89f-94e4-1238-e396-8fb0a0111f97@kernel.dk>
Date: Thu, 11 Oct 2018 15:14:07 -0600
MIME-Version: 1.0
In-Reply-To: <CAFLxGvy04s++HExKixU02=e86kx4yqcz+boeepxi9=uDU94sQg@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
List-ID: <linux-block@vger.kernel.org>

On 10/11/18 3:03 PM, Richard Weinberger wrote:
> Jens,
> 
> On Thu, Oct 11, 2018 at 7:00 PM Jens Axboe <axboe@kernel.dk> wrote:
>>
>> Straight forward conversion, using an internal list to enable the
>> driver to pull requests at will.
>>
>> Dynamically allocate the tag set to avoid having to pull in the
>> block headers for blktrans.h, since various mtd drivers use
>> block conflicting names for defines and functions.
> 
> This explodes on my test system. :-/

I think I see it, that was pretty stupid... Can you try with this one
on top as well?

diff --git a/drivers/mtd/mtd_blkdevs.c b/drivers/mtd/mtd_blkdevs.c
index c26d692781af..e8f8fddce063 100644
--- a/drivers/mtd/mtd_blkdevs.c
+++ b/drivers/mtd/mtd_blkdevs.c
@@ -187,8 +187,8 @@ static void mtd_blktrans_work(struct mtd_blktrans_dev *dev)
 		mutex_unlock(&dev->lock);
 
 		if (!blk_update_request(req, res, blk_rq_cur_bytes(req))) {
-			req = NULL;
 			__blk_mq_end_request(req, res);
+			req = NULL;
 		}
 
 		background_done = 0;

-- 
Jens Axboe