From: Andrew Lunn <andrew@lunn.ch>
To: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: linux-arm-kernel@lists.infradead.org,
linaro-mm-sig@lists.linaro.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org,
Kyungmin Park <kyungmin.park@samsung.com>,
Arnd Bergmann <arnd@arndb.de>, Soren Moch <smoch@web.de>,
Thomas Petazzoni <thomas.petazzoni@free-electrons.com>,
Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>,
Andrew Lunn <andrew@lunn.ch>
Subject: Re: [PATCH] mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls
Date: Sun, 11 Nov 2012 18:22:43 +0100 [thread overview]
Message-ID: <20121111172243.GB821@lunn.ch> (raw)
In-Reply-To: <1352356737-14413-1-git-send-email-m.szyprowski@samsung.com>
On Thu, Nov 08, 2012 at 07:38:57AM +0100, Marek Szyprowski wrote:
> dmapool always calls dma_alloc_coherent() with GFP_ATOMIC flag, regardless
> the flags provided by the caller. This causes excessive pruning of
> emergency memory pools without any good reason. This patch changes the code
> to correctly use gfp flags provided by the dmapool caller. This should
> solve the dmapool usage on ARM architecture, where GFP_ATOMIC DMA
> allocations can be served only from the special, very limited memory pool.
>
> Reported-by: Soren Moch <smoch@web.de>
> Reported-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Andrew Lunn <andrew@lunn.ch>
I tested this on a Kirkwood QNAP after removing the call to
init_dma_coherent_pool_size().
Andrew
> ---
> mm/dmapool.c | 27 +++++++--------------------
> 1 file changed, 7 insertions(+), 20 deletions(-)
>
> diff --git a/mm/dmapool.c b/mm/dmapool.c
> index c5ab33b..86de9b2 100644
> --- a/mm/dmapool.c
> +++ b/mm/dmapool.c
> @@ -62,8 +62,6 @@ struct dma_page { /* cacheable header for 'allocation' bytes */
> unsigned int offset;
> };
>
> -#define POOL_TIMEOUT_JIFFIES ((100 /* msec */ * HZ) / 1000)
> -
> static DEFINE_MUTEX(pools_lock);
>
> static ssize_t
> @@ -227,7 +225,6 @@ static struct dma_page *pool_alloc_page(struct dma_pool *pool, gfp_t mem_flags)
> memset(page->vaddr, POOL_POISON_FREED, pool->allocation);
> #endif
> pool_initialise_page(pool, page);
> - list_add(&page->page_list, &pool->page_list);
> page->in_use = 0;
> page->offset = 0;
> } else {
> @@ -315,30 +312,21 @@ void *dma_pool_alloc(struct dma_pool *pool, gfp_t mem_flags,
> might_sleep_if(mem_flags & __GFP_WAIT);
>
> spin_lock_irqsave(&pool->lock, flags);
> - restart:
> list_for_each_entry(page, &pool->page_list, page_list) {
> if (page->offset < pool->allocation)
> goto ready;
> }
> - page = pool_alloc_page(pool, GFP_ATOMIC);
> - if (!page) {
> - if (mem_flags & __GFP_WAIT) {
> - DECLARE_WAITQUEUE(wait, current);
>
> - __set_current_state(TASK_UNINTERRUPTIBLE);
> - __add_wait_queue(&pool->waitq, &wait);
> - spin_unlock_irqrestore(&pool->lock, flags);
> + /* pool_alloc_page() might sleep, so temporarily drop &pool->lock */
> + spin_unlock_irqrestore(&pool->lock, flags);
>
> - schedule_timeout(POOL_TIMEOUT_JIFFIES);
> + page = pool_alloc_page(pool, mem_flags);
> + if (!page)
> + return NULL;
>
> - spin_lock_irqsave(&pool->lock, flags);
> - __remove_wait_queue(&pool->waitq, &wait);
> - goto restart;
> - }
> - retval = NULL;
> - goto done;
> - }
> + spin_lock_irqsave(&pool->lock, flags);
>
> + list_add(&page->page_list, &pool->page_list);
> ready:
> page->in_use++;
> offset = page->offset;
> @@ -348,7 +336,6 @@ void *dma_pool_alloc(struct dma_pool *pool, gfp_t mem_flags,
> #ifdef DMAPOOL_DEBUG
> memset(retval, POOL_POISON_ALLOCATED, pool->size);
> #endif
> - done:
> spin_unlock_irqrestore(&pool->lock, flags);
> return retval;
> }
> --
> 1.7.9.5
>
next prev parent reply other threads:[~2012-11-11 17:23 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-11-08 6:38 [PATCH] mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls Marek Szyprowski
2012-11-11 17:22 ` Andrew Lunn [this message]
2012-11-12 9:48 ` Soeren Moch
2012-11-12 10:38 ` Andrew Lunn
2012-11-12 11:03 ` Soeren Moch
2012-11-19 0:18 ` Jason Cooper
2012-11-19 22:48 ` Andrew Morton
2012-11-20 10:48 ` Marek Szyprowski
2012-11-20 19:52 ` Andrew Morton
2012-11-20 14:31 ` [PATCH v2] " Marek Szyprowski
2012-11-20 19:33 ` Andrew Morton
2012-11-20 20:27 ` Jason Cooper
2012-11-21 8:08 ` Marek Szyprowski
2012-11-21 8:36 ` Andrew Morton
2012-11-21 9:20 ` Marek Szyprowski
2012-11-21 19:17 ` Andrew Morton
2012-11-22 12:55 ` Marek Szyprowski
2013-01-14 11:56 ` Soeren Moch
2013-01-15 16:56 ` Jason Cooper
2013-01-15 17:50 ` Greg KH
2013-01-15 20:16 ` Jason Cooper
2013-01-15 21:56 ` Jason Cooper
2013-01-16 0:17 ` Soeren Moch
2013-01-16 2:40 ` Jason Cooper
2013-01-16 3:24 ` Soeren Moch
2013-01-16 8:55 ` Soeren Moch
2013-01-16 15:50 ` [PATCH] ata: sata_mv: fix sg_tbl_pool alignment Jason Cooper
2013-01-16 17:05 ` Soeren Moch
2013-01-16 17:52 ` Jason Cooper
2013-01-16 18:35 ` Jason Cooper
2013-01-16 22:26 ` Soeren Moch
2013-01-16 23:10 ` Soeren Moch
2013-01-17 9:11 ` Soeren Moch
2013-01-16 17:32 ` [PATCH v2] mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls Soeren Moch
2013-01-16 17:47 ` Jason Cooper
2013-01-16 22:36 ` Soeren Moch
2013-01-17 10:49 ` Arnd Bergmann
2013-01-17 13:47 ` Soeren Moch
2013-01-17 20:26 ` Arnd Bergmann
2013-01-19 15:29 ` Soeren Moch
2013-01-19 18:59 ` Andrew Lunn
2013-01-23 15:30 ` Soeren Moch
2013-01-23 16:25 ` Andrew Lunn
2013-01-23 17:07 ` Soeren Moch
2013-01-23 17:20 ` Soeren Moch
2013-01-23 18:10 ` Andrew Lunn
2013-01-28 20:59 ` Soeren Moch
2013-01-29 0:13 ` Jason Cooper
2013-01-29 11:02 ` Andrew Lunn
2013-01-29 11:50 ` Soeren Moch
2013-01-19 20:05 ` Arnd Bergmann
2013-01-21 15:01 ` Soeren Moch
2013-01-21 18:55 ` Arnd Bergmann
2013-01-21 21:01 ` Greg KH
2013-01-22 18:13 ` Arnd Bergmann
2013-01-23 14:37 ` Soeren Moch
2013-01-19 16:24 ` Andrew Lunn
2013-01-15 20:05 ` Sebastian Hesselbarth
2013-01-15 20:19 ` Jason Cooper
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20121111172243.GB821@lunn.ch \
--to=andrew@lunn.ch \
--cc=arnd@arndb.de \
--cc=kyungmin.park@samsung.com \
--cc=linaro-mm-sig@lists.linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=m.szyprowski@samsung.com \
--cc=sebastian.hesselbarth@gmail.com \
--cc=smoch@web.de \
--cc=thomas.petazzoni@free-electrons.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).