From: Ilya Yanok <yanok@emcraft.com>
To: linuxppc-dev@lists.ozlabs.org, wd@denx.de, dzu@denx.de, vlad@emcraft.com
Cc: Piotr Ziecik <kosmo@semihalf.com>, Ilya Yanok <yanok@emcraft.com>
Subject: [PATCH 4/6] mpc512x_dma: try to free descriptors in case of allocation failure
Date: Wed, 27 Oct 2010 01:52:58 +0200 [thread overview]
Message-ID: <1288137180-3220-5-git-send-email-yanok@emcraft.com> (raw)
In-Reply-To: <1285676696-5358-1-git-send-email-yanok@emcraft.com>
Currently completed descriptors are processed in the tasklet. This can
lead to dead lock in case of CONFIG_NET_DMA enabled (new requests are
submitted from softirq context and dma_memcpy_to_iovec() busy loops until
the requests is submitted). To prevent this we should process completed
descriptors from the allocation failure path in prepare_memcpy too.
Signed-off-by: Ilya Yanok <yanok@emcraft.com>
Cc: Piotr Ziecik <kosmo@semihalf.com>
---
drivers/dma/mpc512x_dma.c | 79 +++++++++++++++++++++++++-------------------
1 files changed, 45 insertions(+), 34 deletions(-)
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 97b92ec..59c2701 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -328,19 +328,55 @@ static irqreturn_t mpc_dma_irq(int irq, void *data)
return IRQ_HANDLED;
}
-/* DMA Tasklet */
-static void mpc_dma_tasklet(unsigned long data)
+/* proccess completed descriptors */
+static void mpc_dma_process_completed(struct mpc_dma *mdma)
{
- struct mpc_dma *mdma = (void *)data;
dma_cookie_t last_cookie = 0;
struct mpc_dma_chan *mchan;
struct mpc_dma_desc *mdesc;
struct dma_async_tx_descriptor *desc;
unsigned long flags;
LIST_HEAD(list);
- uint es;
int i;
+ for (i = 0; i < mdma->dma.chancnt; i++) {
+ mchan = &mdma->channels[i];
+
+ /* Get all completed descriptors */
+ spin_lock_irqsave(&mchan->lock, flags);
+ if (!list_empty(&mchan->completed))
+ list_splice_tail_init(&mchan->completed, &list);
+ spin_unlock_irqrestore(&mchan->lock, flags);
+
+ if (list_empty(&list))
+ continue;
+
+ /* Execute callbacks and run dependencies */
+ list_for_each_entry(mdesc, &list, node) {
+ desc = &mdesc->desc;
+
+ if (desc->callback)
+ desc->callback(desc->callback_param);
+
+ last_cookie = desc->cookie;
+ dma_run_dependencies(desc);
+ }
+
+ /* Free descriptors */
+ spin_lock_irqsave(&mchan->lock, flags);
+ list_splice_tail_init(&list, &mchan->free);
+ mchan->completed_cookie = last_cookie;
+ spin_unlock_irqrestore(&mchan->lock, flags);
+ }
+}
+
+/* DMA Tasklet */
+static void mpc_dma_tasklet(unsigned long data)
+{
+ struct mpc_dma *mdma = (void *)data;
+ unsigned long flags;
+ uint es;
+
spin_lock_irqsave(&mdma->error_status_lock, flags);
es = mdma->error_status;
mdma->error_status = 0;
@@ -379,35 +415,7 @@ static void mpc_dma_tasklet(unsigned long data)
dev_err(mdma->dma.dev, "- Destination Bus Error\n");
}
- for (i = 0; i < mdma->dma.chancnt; i++) {
- mchan = &mdma->channels[i];
-
- /* Get all completed descriptors */
- spin_lock_irqsave(&mchan->lock, flags);
- if (!list_empty(&mchan->completed))
- list_splice_tail_init(&mchan->completed, &list);
- spin_unlock_irqrestore(&mchan->lock, flags);
-
- if (list_empty(&list))
- continue;
-
- /* Execute callbacks and run dependencies */
- list_for_each_entry(mdesc, &list, node) {
- desc = &mdesc->desc;
-
- if (desc->callback)
- desc->callback(desc->callback_param);
-
- last_cookie = desc->cookie;
- dma_run_dependencies(desc);
- }
-
- /* Free descriptors */
- spin_lock_irqsave(&mchan->lock, flags);
- list_splice_tail_init(&list, &mchan->free);
- mchan->completed_cookie = last_cookie;
- spin_unlock_irqrestore(&mchan->lock, flags);
- }
+ mpc_dma_process_completed(mdma);
}
/* Submit descriptor to hardware */
@@ -587,8 +595,11 @@ mpc_dma_prep_memcpy(struct dma_chan *chan, dma_addr_t dst, dma_addr_t src,
}
spin_unlock_irqrestore(&mchan->lock, iflags);
- if (!mdesc)
+ if (!mdesc) {
+ /* try to free completed descriptors */
+ mpc_dma_process_completed(mdma);
return NULL;
+ }
mdesc->error = 0;
tcd = mdesc->tcd;
--
1.7.2.3
next prev parent reply other threads:[~2010-10-26 23:53 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-09-28 12:24 [RFC][PATCH 0/3] fixes and MPC8308 support for the mpc512x_dma driver Ilya Yanok
2010-09-28 12:24 ` [PATCH 1/3] mpc512x_dma: scatter/gather fix Ilya Yanok
2010-09-28 12:24 ` [PATCH 2/3] mpc512x_dma: fix the hanged transfer issue Ilya Yanok
2010-09-28 12:24 ` [PATCH 3/3] mpc512x_dma: add MPC8308 support Ilya Yanok
2010-09-28 13:09 ` Wolfgang Denk
2010-09-28 13:47 ` Ilya Yanok
2010-10-26 23:52 ` [REPOST] [PATCH 0/6] fixes and MPC8308 support for the mpc512x_dma driver Ilya Yanok
2010-10-27 7:24 ` Piotr Zięcik
2010-10-28 0:44 ` Ilya Yanok
2010-11-11 12:11 ` Wolfgang Denk
2010-10-26 23:52 ` [PATCH 1/6] mpc512x_dma: scatter/gather fix Ilya Yanok
2010-12-30 5:35 ` Grant Likely
2010-10-26 23:52 ` [PATCH 2/6] mpc512x_dma: fix the hanged transfer issue Ilya Yanok
2010-12-30 5:36 ` Grant Likely
2010-10-26 23:52 ` [PATCH 3/6] mpc512x_dma: add MPC8308 support Ilya Yanok
2010-12-30 5:35 ` Grant Likely
2010-10-26 23:52 ` Ilya Yanok [this message]
2010-12-30 5:36 ` [PATCH 4/6] mpc512x_dma: try to free descriptors in case of allocation failure Grant Likely
2010-10-26 23:52 ` [PATCH 5/6] MPC8308RDB: add DMA controller device-tree node Ilya Yanok
2010-12-30 5:36 ` Grant Likely
2010-10-26 23:53 ` [PATCH 6/6] mpc8308_p1m: " Ilya Yanok
2010-12-30 5:35 ` Grant Likely
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=1288137180-3220-5-git-send-email-yanok@emcraft.com \
--to=yanok@emcraft.com \
--cc=dzu@denx.de \
--cc=kosmo@semihalf.com \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=vlad@emcraft.com \
--cc=wd@denx.de \
/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).