linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Tudor Ambarus <tudor.ambarus@microchip.com>
To: <vkoul@kernel.org>, <peda@axentia.se>, <du@axentia.se>
Cc: <maciej.sosnowski@intel.com>, <nicolas.ferre@microchip.com>,
	<mripard@kernel.org>, <torfl6749@gmail.com>,
	<linux-kernel@vger.kernel.org>, <dmaengine@vger.kernel.org>,
	<linux-arm-kernel@lists.infradead.org>,
	"Tudor Ambarus" <tudor.ambarus@microchip.com>,
	<stable@vger.kernel.org>
Subject: [PATCH v2 07/32] dmaengine: at_hdmac: Fix concurrency problems by removing atc_complete_all()
Date: Tue, 25 Oct 2022 12:02:41 +0300	[thread overview]
Message-ID: <20221025090306.297886-8-tudor.ambarus@microchip.com> (raw)
In-Reply-To: <20221025090306.297886-1-tudor.ambarus@microchip.com>

atc_complete_all() had concurrency bugs, thus remove it:
1/ atc_complete_all() in its entirety was buggy, as when the atchan->queue
list (the one that contains descriptors that are not yet issued to the
hardware) contained descriptors, it fired just the first from the
atchan->queue, but moved all the desc from atchan->queue to
atchan->active_list and considered them all as fired. This could result in
calling the completion of a descriptor that was not yet issued to the
hardware.
2/ when in tasklet at atc_advance_work() time, atchan->active_list was
queried without holding the lock of the chan. This can result in
atchan->active_list concurrency problems between the tasklet and
issue_pending().

Fixes: dc78baa2b90b ("dmaengine: at_hdmac: new driver for the Atmel AHB DMA Controller")
Reported-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Tudor Ambarus <tudor.ambarus@microchip.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/lkml/13c6c9a2-6db5-c3bf-349b-4c127ad3496a@axentia.se/
---
 drivers/dma/at_hdmac.c | 49 ++++--------------------------------------
 1 file changed, 4 insertions(+), 45 deletions(-)

diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index deb4c6027436..f1e6fa6af6c2 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -485,42 +485,6 @@ atc_chain_complete(struct at_dma_chan *atchan, struct at_desc *desc)
 	dma_run_dependencies(txd);
 }
 
-/**
- * atc_complete_all - finish work for all transactions
- * @atchan: channel to complete transactions for
- *
- * Eventually submit queued descriptors if any
- *
- * Assume channel is idle while calling this function
- * Called with atchan->lock held and bh disabled
- */
-static void atc_complete_all(struct at_dma_chan *atchan)
-{
-	struct at_desc *desc, *_desc;
-	LIST_HEAD(list);
-	unsigned long flags;
-
-	dev_vdbg(chan2dev(&atchan->chan_common), "complete all\n");
-
-	spin_lock_irqsave(&atchan->lock, flags);
-
-	/*
-	 * Submit queued descriptors ASAP, i.e. before we go through
-	 * the completed ones.
-	 */
-	if (!list_empty(&atchan->queue))
-		atc_dostart(atchan, atc_first_queued(atchan));
-	/* empty active_list now it is completed */
-	list_splice_init(&atchan->active_list, &list);
-	/* empty queue list by moving descriptors (if any) to active_list */
-	list_splice_init(&atchan->queue, &atchan->active_list);
-
-	spin_unlock_irqrestore(&atchan->lock, flags);
-
-	list_for_each_entry_safe(desc, _desc, &list, desc_node)
-		atc_chain_complete(atchan, desc);
-}
-
 /**
  * atc_advance_work - at the end of a transaction, move forward
  * @atchan: channel where the transaction ended
@@ -528,25 +492,20 @@ static void atc_complete_all(struct at_dma_chan *atchan)
 static void atc_advance_work(struct at_dma_chan *atchan)
 {
 	unsigned long flags;
-	int ret;
 
 	dev_vdbg(chan2dev(&atchan->chan_common), "advance_work\n");
 
 	spin_lock_irqsave(&atchan->lock, flags);
-	ret = atc_chan_is_enabled(atchan);
+	if (atc_chan_is_enabled(atchan) || list_empty(&atchan->active_list))
+		return spin_unlock_irqrestore(&atchan->lock, flags);
 	spin_unlock_irqrestore(&atchan->lock, flags);
-	if (ret)
-		return;
-
-	if (list_empty(&atchan->active_list) ||
-	    list_is_singular(&atchan->active_list))
-		return atc_complete_all(atchan);
 
 	atc_chain_complete(atchan, atc_first_active(atchan));
 
 	/* advance work */
 	spin_lock_irqsave(&atchan->lock, flags);
-	atc_dostart(atchan, atc_first_active(atchan));
+	if (!list_empty(&atchan->active_list))
+		atc_dostart(atchan, atc_first_active(atchan));
 	spin_unlock_irqrestore(&atchan->lock, flags);
 }
 
-- 
2.25.1


  parent reply	other threads:[~2022-10-25  9:04 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-25  9:02 [PATCH v2 00/32] dmaengine: at_hdmac: Fix concurrency bugs and then convert to virt-dma Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 01/32] dmaengine: at_hdmac: Fix at_lli struct definition Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 02/32] dmaengine: at_hdmac: Don't start transactions at tx_submit level Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 03/32] dmaengine: at_hdmac: Start transfer for cyclic channels in issue_pending Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 04/32] dmaengine: at_hdmac: Fix premature completion of desc " Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 05/32] dmaengine: at_hdmac: Do not call the complete callback on device_terminate_all Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 06/32] dmaengine: at_hdmac: Protect atchan->status with the channel lock Tudor Ambarus
2022-10-25  9:02 ` Tudor Ambarus [this message]
2022-10-25  9:02 ` [PATCH v2 08/32] dmaengine: at_hdmac: Fix concurrency over descriptor Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 09/32] dmaengine: at_hdmac: Free the memset buf without holding the chan lock Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 10/32] dmaengine: at_hdmac: Fix concurrency over the active list Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 11/32] dmaengine: at_hdmac: Fix descriptor handling when issuing it to hardware Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 12/32] dmaengine: at_hdmac: Fix completion of unissued descriptor in case of errors Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 13/32] dmaengine: at_hdmac: Don't allow CPU to reorder channel enable Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 14/32] dmaengine: at_hdmac: Fix impossible condition Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 15/32] dmaengine: at_hdmac: Check return code of dma_async_device_register Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 16/32] dmaengine: at_hdmac: Do not print messages on console while holding the lock Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 17/32] dmaengine: at_hdmac: Return dma_cookie_status()'s ret code when txstate is NULL Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 18/32] dmaengine: at_hdmac: Remove superfluous cast Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 19/32] dmaengine: at_hdmac: Pass residue by address to avoid unnecessary implicit casts Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 20/32] dmaengine: at_hdmac: s/atc_get_bytes_left/atc_get_residue Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 21/32] dmaengine: at_hdmac: Introduce atc_get_llis_residue() Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 22/32] dmaengine: at_hdmac: Use devm_kzalloc() and struct_size() Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 23/32] dmaengine: at_hdmac: Use devm_platform_ioremap_resource Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 24/32] dmaengine: at_hdmac: Use devm_clk_get() Tudor Ambarus
2022-10-25  9:02 ` [PATCH v2 25/32] dmaengine: at_hdmac: Use pm_ptr() Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 26/32] dmaengine: at_hdmac: Set include entries in alphabetic order Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 27/32] dmaengine: at_hdmac: Keep register definitions and structures private to at_hdmac.c Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 28/32] dmaengine: at_hdmac: Use bitfield access macros Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 29/32] dmaengine: at_hdmac: Rename "dma_common" to "dma_device" Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 30/32] dmaengine: at_hdmac: Rename "chan_common" to "dma_chan" Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 31/32] dmaengine: at_hdmac: Remove unused member of at_dma_chan Tudor Ambarus
2022-10-25  9:03 ` [PATCH v2 32/32] dmaengine: at_hdmac: Convert driver to use virt-dma Tudor Ambarus
2022-10-28 12:55   ` Nicolas Ferre
2022-10-25  9:06 ` [PATCH v2 00/32] dmaengine: at_hdmac: Fix concurrency bugs and then convert to virt-dma Tudor.Ambarus
2022-10-28 13:00 ` Nicolas Ferre
2022-11-04 13:22 ` Vinod Koul
2022-11-04 15:14   ` Tudor.Ambarus
2022-11-04 15:21     ` Tudor.Ambarus
2022-11-04 17:59       ` Vinod Koul
2022-11-07  4:33         ` Vinod Koul
2022-11-07  9:00           ` Tudor.Ambarus

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=20221025090306.297886-8-tudor.ambarus@microchip.com \
    --to=tudor.ambarus@microchip.com \
    --cc=dmaengine@vger.kernel.org \
    --cc=du@axentia.se \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=maciej.sosnowski@intel.com \
    --cc=mripard@kernel.org \
    --cc=nicolas.ferre@microchip.com \
    --cc=peda@axentia.se \
    --cc=stable@vger.kernel.org \
    --cc=torfl6749@gmail.com \
    --cc=vkoul@kernel.org \
    /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).