From: Tudor Ambarus <tudor.ambarus@microchip.com>
To: <vkoul@kernel.org>
Cc: Tudor Ambarus <tudor.ambarus@microchip.com>,
mripard@kernel.org, linux-kernel@vger.kernel.org,
ludovic.desroches@microchip.com, dmaengine@vger.kernel.org,
linux-arm-kernel@lists.infradead.org
Subject: [PATCH v3 12/12] dmaengine: at_xdmac: Fix race over irq_status
Date: Wed, 15 Dec 2021 13:01:15 +0200 [thread overview]
Message-ID: <20211215110115.191749-13-tudor.ambarus@microchip.com> (raw)
In-Reply-To: <20211215110115.191749-1-tudor.ambarus@microchip.com>
Tasklets run with interrupts enabled, so we need to protect
atchan->irq_status with spin_lock_irq() otherwise the tasklet can be
interrupted by the IRQ that modifies irq_status. Move the dev_dbg that
prints the irq_status in at_xdmac_handle_cyclic() and lower in
at_xdmac_tasklet() where the IRQ is disabled.
Signed-off-by: Tudor Ambarus <tudor.ambarus@microchip.com>
---
drivers/dma/at_xdmac.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c
index ba727751a9f6..a1da2b4b6d73 100644
--- a/drivers/dma/at_xdmac.c
+++ b/drivers/dma/at_xdmac.c
@@ -1611,6 +1611,8 @@ static void at_xdmac_handle_cyclic(struct at_xdmac_chan *atchan)
struct dma_async_tx_descriptor *txd;
spin_lock_irq(&atchan->lock);
+ dev_dbg(chan2dev(&atchan->chan), "%s: status=0x%08x\n",
+ __func__, atchan->irq_status);
if (list_empty(&atchan->xfers_list)) {
spin_unlock_irq(&atchan->lock);
return;
@@ -1623,6 +1625,7 @@ static void at_xdmac_handle_cyclic(struct at_xdmac_chan *atchan)
dmaengine_desc_get_callback_invoke(txd, NULL);
}
+/* Called with atchan->lock held. */
static void at_xdmac_handle_error(struct at_xdmac_chan *atchan)
{
struct at_xdmac *atxdmac = to_at_xdmac(atchan->chan.device);
@@ -1641,8 +1644,6 @@ static void at_xdmac_handle_error(struct at_xdmac_chan *atchan)
if (atchan->irq_status & AT_XDMAC_CIS_ROIS)
dev_err(chan2dev(&atchan->chan), "request overflow error!!!");
- spin_lock_irq(&atchan->lock);
-
/* Channel must be disabled first as it's not done automatically */
at_xdmac_write(atxdmac, AT_XDMAC_GD, atchan->mask);
while (at_xdmac_read(atxdmac, AT_XDMAC_GS) & atchan->mask)
@@ -1652,8 +1653,6 @@ static void at_xdmac_handle_error(struct at_xdmac_chan *atchan)
struct at_xdmac_desc,
xfer_node);
- spin_unlock_irq(&atchan->lock);
-
/* Print bad descriptor's details if needed */
dev_dbg(chan2dev(&atchan->chan),
"%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x\n",
@@ -1670,15 +1669,17 @@ static void at_xdmac_tasklet(struct tasklet_struct *t)
struct dma_async_tx_descriptor *txd;
u32 error_mask;
- dev_dbg(chan2dev(&atchan->chan), "%s: status=0x%08x\n",
- __func__, atchan->irq_status);
-
if (at_xdmac_chan_is_cyclic(atchan))
return at_xdmac_handle_cyclic(atchan);
error_mask = AT_XDMAC_CIS_RBEIS | AT_XDMAC_CIS_WBEIS |
AT_XDMAC_CIS_ROIS;
+ spin_lock_irq(&atchan->lock);
+
+ dev_dbg(chan2dev(&atchan->chan), "%s: status=0x%08x\n",
+ __func__, atchan->irq_status);
+
if (!(atchan->irq_status & AT_XDMAC_CIS_LIS) &&
!(atchan->irq_status & error_mask))
return;
@@ -1686,7 +1687,6 @@ static void at_xdmac_tasklet(struct tasklet_struct *t)
if (atchan->irq_status & error_mask)
at_xdmac_handle_error(atchan);
- spin_lock_irq(&atchan->lock);
desc = list_first_entry(&atchan->xfers_list, struct at_xdmac_desc,
xfer_node);
dev_vdbg(chan2dev(&atchan->chan), "%s: desc 0x%p\n", __func__, desc);
--
2.25.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2021-12-15 11:17 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-15 11:01 [PATCH v3 00/12] dmaengine: at_xdmac: Various fixes Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 01/12] dmaengine: at_xdmac: Don't start transactions at tx_submit level Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 02/12] dmaengine: at_xdmac: Start transfer for cyclic channels in issue_pending Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 03/12] dmaengine: at_xdmac: Print debug message after realeasing the lock Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 04/12] dmaengine: at_xdmac: Fix concurrency over chan's completed_cookie Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 05/12] dmaengine: at_xdmac: Fix race for the tx desc callback Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 06/12] dmaengine: at_xdmac: Move the free desc to the tail of the desc list Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 07/12] dmaengine: at_xdmac: Fix concurrency over xfers_list Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 08/12] dmaengine: at_xdmac: Remove a level of indentation in at_xdmac_advance_work() Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 09/12] dmaengine: at_xdmac: Fix lld view setting Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 10/12] dmaengine: at_xdmac: Fix at_xdmac_lld struct definition Tudor Ambarus
2021-12-15 11:01 ` [PATCH v3 11/12] dmaengine: at_xdmac: Remove a level of indentation in at_xdmac_tasklet() Tudor Ambarus
2021-12-15 11:01 ` Tudor Ambarus [this message]
2022-01-05 10:20 ` [PATCH v3 00/12] dmaengine: at_xdmac: Various fixes Vinod Koul
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=20211215110115.191749-13-tudor.ambarus@microchip.com \
--to=tudor.ambarus@microchip.com \
--cc=dmaengine@vger.kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ludovic.desroches@microchip.com \
--cc=mripard@kernel.org \
--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