2018-04-10 01:50:51

by Jia-Ju Bai

Subject: [PATCH] video: fbdev: aty: radeon_pm: Replace mdelay with msleep in radeonfb_pci_suspend

radeonfb_pci_suspend() is never called in atomic context.

radeonfb_pci_suspend() is only set as ".suspend" in struct pci_driver.
This function is not called in atomic context.

Despite never getting called from atomic context, radeonfb_pci_suspend()
calls mdelay() to busily wait.
This is not necessary and can be replaced with msleep()
and usleep_range() to avoid busy waiting.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <[email protected]>
---
drivers/video/fbdev/aty/radeon_pm.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/video/fbdev/aty/radeon_pm.c b/drivers/video/fbdev/aty/radeon_pm.c
index 1417542..a8de909 100644
--- a/drivers/video/fbdev/aty/radeon_pm.c
+++ b/drivers/video/fbdev/aty/radeon_pm.c
@@ -2674,17 +2674,17 @@ int radeonfb_pci_suspend(struct pci_dev *pdev, pm_message_t mesg)
* it, we'll restore the dynamic clocks state on wakeup
*/
radeon_pm_disable_dynamic_mode(rinfo);
- mdelay(50);
+ msleep(50);
radeon_pm_save_regs(rinfo, 1);

if (rinfo->is_mobility && !(rinfo->pm_mode & radeon_pm_d2)) {
/* Switch off LVDS interface */
- mdelay(1);
+ usleep_range(1000, 2000);
OUTREG(LVDS_GEN_CNTL, INREG(LVDS_GEN_CNTL) & ~(LVDS_BL_MOD_EN));
- mdelay(1);
+ usleep_range(1000, 2000);
OUTREG(LVDS_GEN_CNTL, INREG(LVDS_GEN_CNTL) & ~(LVDS_EN | LVDS_ON));
OUTREG(LVDS_PLL_CNTL, (INREG(LVDS_PLL_CNTL) & ~30000) | 0x20000);
- mdelay(20);
+ msleep(20);
OUTREG(LVDS_GEN_CNTL, INREG(LVDS_GEN_CNTL) & ~(LVDS_DIGON));
}
pci_disable_device(pdev);
--
1.9.1

2018-04-24 15:53:31

by Bartlomiej Zolnierkiewicz

Subject: Re: [PATCH] video: fbdev: aty: radeon_pm: Replace mdelay with msleep in radeonfb_pci_suspend

On Tuesday, April 10, 2018 09:47:20 AM Jia-Ju Bai wrote:
> radeonfb_pci_suspend() is never called in atomic context.
>
> radeonfb_pci_suspend() is only set as ".suspend" in struct pci_driver.
> This function is not called in atomic context.
>
> Despite never getting called from atomic context, radeonfb_pci_suspend()
> calls mdelay() to busily wait.
> This is not necessary and can be replaced with msleep()
> and usleep_range() to avoid busy waiting.
>
> This is found by a static analysis tool named DCNS written by myself.
> And I also manually check it.
>
> Signed-off-by: Jia-Ju Bai <[email protected]>

Patch queued for 4.18, thanks.

Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics