at_hdmac: check and return DMA_PAUSED status when suitable
Subsystems (2)
PATCH 1/2 at_hdmac: check and return DMA_PAUSED status when suitable
device_tx_status() may return DMA_PAUSED status when driver supports it.
--- drivers/dma/at_hdmac.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
drivers/dma/at_hdmac.c +14 -1
@@ -1669,9 +1669,19 @@ atc_tx_status(struct dma_chan *chan,int ret; dma_status =dma_cookie_status (chan, cookie, txstate); -if (dma_status ==DMA_COMPLETE || !txstate) +if (dma_status ==DMA_COMPLETE )return dma_status; +/* +* There's no point in calculating the residue if there's +* no txstate to store the value. +*/ +if (!txstate) { +if (test_bit (ATC_IS_PAUSED , &atchan->status)) +return DMA_PAUSED ; +return DMA_ERROR ; + } +spin_lock_irqsave (&atchan->vc.lock, flags);/* Get number of bytes left in the active transactions */ ret =atc_get_residue (chan, cookie, &residue);
@@ -1684,6 +1694,9 @@ atc_tx_status(struct dma_chan *chan,dma_set_residue (txstate, residue); } +if (test_bit (ATC_IS_PAUSED , &atchan->status)) + dma_status =DMA_PAUSED ; +dev_vdbg (chan2dev (chan),"tx_status %d: cookie = %d residue = %u\n" , dma_status, cookie, residue);
Signature · 1 line
2.35.1
PATCH 2/2 at_hdmac: return actual status when txstatus parameter is NULL
There is no point to return DMA_ERROR if txstatus parameter is NULL. It's a valid case and should be handled correspondingly.
--- drivers/dma/at_hdmac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/at_hdmac.c +1 -1
@@ -1679,7 +1679,7 @@ atc_tx_status(struct dma_chan *chan,if (!txstate) {if (test_bit (ATC_IS_PAUSED , &atchan->status))return DMA_PAUSED ; -return DMA_ERROR ; +return dma_status; }spin_lock_irqsave (&atchan->vc.lock, flags);
Signature · 1 line
2.35.1
Hi, Andy,
Quoted 3 lines · Andy Shevchenko
EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
device_tx_status()may returnDMA_PAUSEDstatus when driver supports it.
Yeah, but you haven't described the why. And that's because dma_cookie_status() currently considers just DMA_COMPLETE and DMA_IN_PROGRESS for the status of the cookie, so the controller drivers are forced to query the DMA_PAUSED state themselves.
Also, I noticed that Vinod prefers that you use the full paragraph in the commit message, and not a continuation of the patch title, so you may want to reword the commit message.
Quoted 14 lines
Signed-off-by Andy Shevchenko <[email protected]>--- drivers/dma/at_hdmac.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)drivers/dma/at_hdmac.c +0
-0@@ -1669,9 +1669,19 @@ atc_tx_status(struct dma_chan *chan,int ret; dma_status =dma_cookie_status (chan, cookie, txstate);
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1680
Quoted 10 lines
drivers/dma/at_hdmac.c +7
-1-if (dma_status ==DMA_COMPLETE || !txstate) +if (dma_status ==DMA_COMPLETE )return dma_status; +/* +* There's no point in calculating the residue if there's +* no txstate to store the value. +*/ +if (!txstate) { +if (test_bit (ATC_IS_PAUSED , &atchan->status))
there's a helper function that you can use instead: atc_chan_is_paused()
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1682
Quoted 2 lines
drivers/dma/at_hdmac.c +2
-0+return DMA_PAUSED ; +return DMA_ERROR ;
return dma_status please
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1697
Quoted 10 lines
drivers/dma/at_hdmac.c +3
-0+ } +spin_lock_irqsave (&atchan->vc.lock, flags);/* Get number of bytes left in the active transactions */ ret =atc_get_residue (chan, cookie, &residue);@@ -1684,6 +1694,9 @@ atc_tx_status(struct dma_chan *chan,dma_set_residue (txstate, residue); } +if (test_bit (ATC_IS_PAUSED , &atchan->status))
The status may change after spin_unlock_irqrestore(). Should the residue be in sync with the dma status? If yes, you should check the status while holding the lock.
Quoted 8 lines
drivers/dma/at_hdmac.c +2
-1+ dma_status =DMA_PAUSED ; +dev_vdbg (chan2dev (chan),"tx_status %d: cookie = %d residue = %u\n" , dma_status, cookie, residue); --2.35.1
Signature · 2 lines
Cheers, ta
Re: [PATCH v1 2/2] at_hdmac: return actual status when txstatus parameter is NULL
In reply to Andy ShevchenkoQuoted 15 lines · Andy Shevchenko
EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
There is no point to return
DMA_ERRORif txstatus parameter is NULL. It's a valid case and should be handled correspondingly.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/dma/at_hdmac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/dma/at_hdmac.c +0
-0
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1682
Quoted 5 lines
drivers/dma/at_hdmac.c +1
-1@@ -1679,7 +1679,7 @@ atc_tx_status(struct dma_chan *chan,if (!txstate) {if (test_bit (ATC_IS_PAUSED , &atchan->status))return DMA_PAUSED ; -return DMA_ERROR ; +return dma_status;
oh, I see you do it here, please squash this patch to the previous one, you introduced DMA_ERROR in the previous patch.
Quoted 6 lines
drivers/dma/at_hdmac.c +0
-1}spin_lock_irqsave (&atchan->vc.lock, flags); --2.35.1
Signature · 2 lines
Cheers, ta
Quoted 9 lines · On Tue, Nov 08, 2022 at 11:50:19AM +0000, [email protected] wrote:
Quoted 3 lines · Andy Shevchenko
EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
device_tx_status()may returnDMA_PAUSEDstatus when driver supports it.Yeah, but you haven't described the why. And that's because
dma_cookie_status()currently considers justDMA_COMPLETEandDMA_IN_PROGRESSfor the status of the cookie, so the controller drivers are forced to query theDMA_PAUSEDstate themselves.
At last without this change it's inconvenient and requires a lot of additional (unneeded) code to be written by the caller. Moreover, it's racy. If you query status twice in a raw, it may be well changed (for example from PAUSED to IN_PROGRESS or COMPETE).
I will add a word summarizing this.
Quoted 3 lines
Also, I noticed that Vinod prefers that you use the full paragraph in the commit message, and not a continuation of the patch title, so you may want to reword the commit message.
Yes, I will fix that.
...
Quoted 4 lines
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1680
Quoted 2 lines
drivers/dma/at_hdmac.c +2
-0+if (!txstate) { +if (test_bit (ATC_IS_PAUSED , &atchan->status))there's a helper function that you can use instead:
atc_chan_is_paused()
Will use it.
Quoted 4 lines
drivers/dma/at_hdmac.c atc_tx_status(struct dma_chan *chan, · line 1682
Quoted 2 lines
drivers/dma/at_hdmac.c +2
-0+return DMA_PAUSED ; +return DMA_ERROR ;return
dma_statusplease
Will squash the patch.
Quoted 1 line
Quoted 1 line
+ }
...
Quoted 5 lines
Quoted 1 line
+ if (
test_bit(ATC_IS_PAUSED, &atchan->status))The status may change after
spin_unlock_irqrestore(). Should the residue be in sync with the dma status? If yes, you should check the status while holding the lock.
You should tell me actually. Because I'm a bit puzzled why we need a spin lock _and_ atomic bit operations together.
Quoted 1 line
Quoted 1 line
+
dma_status=DMA_PAUSED;
...
Thank you for the review.
Signature · 2 lines
With Best Regards, Andy Shevchenko