Skip to contentSkip to conversation

dmaengine: uniphier-xdmac: Add UniPhier external DMA controller driver

bug report Dan Carpenter 2 participants 3 messages Jul 26, 2021 – Jul 27, 2021 dmaengine

lore
Share & tools
lore
#0 lore raw

Hello Kunihiko Hayashi,

The patch 667b9251440b: "dmaengine: uniphier-xdmac: Add UniPhier external DMA controller driver" from Feb 21, 2020, leads to the following static checker warning:

drivers/dma/uniphier-xdmac.c:212 uniphier_xdmac_chan_stop()
warn: sleeping in atomic context

drivers/dma/uniphier-xdmac.c

    197 static int uniphier_xdmac_chan_stop(struct uniphier_xdmac_chan *xc)
    198 {
    199 	u32 val;
    200 
    201 	/* disable interrupt */
    202 	val = readl(xc->reg_ch_base + XDMAC_IEN);
    203 	val &= ~(XDMAC_IEN_ENDIEN | XDMAC_IEN_ERRIEN);
    204 	writel(val, xc->reg_ch_base + XDMAC_IEN);
    205 
    206 	/* stop XDMAC */
    207 	val = readl(xc->reg_ch_base + XDMAC_TSS);
    208 	val &= ~XDMAC_TSS_REQ;
    209 	writel(0, xc->reg_ch_base + XDMAC_TSS);
    210 
    211 	/* wait until transfer is stopped */
--> 212 	return readl_poll_timeout(xc->reg_ch_base + XDMAC_STAT, val,
    213 				  !(val & XDMAC_STAT_TENF), 100, 1000);
                                                                    ^^^

This is doing a 100 us sleep but both callers hold a spin lock.

    214 }

regards,
dan carpenter

#1 lore raw

Hi Dan,
Thank you for reporting.

Quoted 28 lines · Dan Carpenter

Hello Kunihiko Hayashi,

The patch 667b9251440b: "dmaengine: uniphier-xdmac: Add UniPhier external DMA controller driver" from Feb 21, 2020, leads to the following static checker warning:

drivers/dma/uniphier-xdmac.c:212 uniphier_xdmac_chan_stop()
warn: sleeping in atomic context

drivers/dma/uniphier-xdmac.c

     197 static int uniphier_xdmac_chan_stop(struct uniphier_xdmac_chan *xc)
     198 {
     199 	u32 val;
     200
     201 	/* disable interrupt */
     202 	val = readl(xc->reg_ch_base + XDMAC_IEN);
     203 	val &= ~(XDMAC_IEN_ENDIEN | XDMAC_IEN_ERRIEN);
     204 	writel(val, xc->reg_ch_base + XDMAC_IEN);
     205
     206 	/* stop XDMAC */
     207 	val = readl(xc->reg_ch_base + XDMAC_TSS);
     208 	val &= ~XDMAC_TSS_REQ;
     209 	writel(0, xc->reg_ch_base + XDMAC_TSS);
     210
     211 	/* wait until transfer is stopped */
--> 212 	return readl_poll_timeout(xc->reg_ch_base + XDMAC_STAT, val,
     213 				  !(val & XDMAC_STAT_TENF), 100, 1000);
                                                                     ^^^
Quoted 1 line

This is doing a 100 us sleep but both callers hold a spin lock.

Yes, this waiting code is in spin lock context.
I think that the function needs to be rewritten in tasklet.

Thank you,

---
Best Regards
Kunihiko Hayashi

#2 lore raw

Hi,

Quoted 34 lines · Kunihiko Hayashi

Hi Dan,
Thank you for reporting.

Quoted 21 lines · Dan Carpenter

Hello Kunihiko Hayashi,

The patch 667b9251440b: "dmaengine: uniphier-xdmac: Add UniPhier external DMA controller driver" from Feb 21, 2020, leads to the following static checker warning:

    drivers/dma/uniphier-xdmac.c:212 uniphier_xdmac_chan_stop()
    warn: sleeping in atomic context

drivers/dma/uniphier-xdmac.c
     197 static int uniphier_xdmac_chan_stop(struct uniphier_xdmac_chan *xc) 198 {
     199     u32 val;
     200
     201     /* disable interrupt */
     202     val = readl(xc->reg_ch_base + XDMAC_IEN);
     203     val &= ~(XDMAC_IEN_ENDIEN | XDMAC_IEN_ERRIEN);
     204     writel(val, xc->reg_ch_base + XDMAC_IEN);
     205
     206     /* stop XDMAC */
     207     val = readl(xc->reg_ch_base + XDMAC_TSS);

Quoted 8 lines

     208     val &= ~XDMAC_TSS_REQ;
     209     writel(0, xc->reg_ch_base + XDMAC_TSS);
     210
     211     /* wait until transfer is stopped */
--> 212     return readl_poll_timeout(xc->reg_ch_base + XDMAC_STAT, val, 213                   !(val & XDMAC_STAT_TENF), 100, 1000); ^^^ This is doing a 100 us sleep but both callers hold a spin lock.

Quoted 2 lines

Yes, this waiting code is in spin lock context.
I think that the function needs to be rewritten in tasklet.

I found this can be simply replaced with readl_poll_timeout_atomic(). I'll fix it.

Thank you,
---
Best Regards
Kunihiko Hayashi