| Message ID | 20200410015832.8012-1-tangbin@cmss.chinamobile.com (mailing list archive) |
|---|---|
| State | Mainlined |
| Commit | 4cda340a455b425f7df9657aaaa78a75757d940d |
| Headers | show |
| Series | usb: gadget: fsl: Fix a wrong judgment in fsl_udc_probe() | expand |
> If the function "platform_get_irq()" failed, the negative value > returned will not be detected here, including "-EPROBE_DEFER", I suggest to adjust this change description. Wording alternative: The negative return value (which could eventually be “-EPROBE_DEFER”) will not be detected here from a failed call of the function “platform_get_irq”. > which causes the application to fail to get the correct error message. Will another fine-tuning become relevant also for this wording? > Thus it must be fixed. Wording alternative: Thus adjust the error detection and corresponding exception handling. > Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com> > Signed-off-by: Shengju Zhang <zhangshengju@cmss.chinamobile.com> How do you think about to add the tags “Fixes”, “Link” and “Reported-by”? https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=c0cc271173b2e1c2d8d0ceaef14e4dfa79eefc0d#n584 usb: gadget: fsl_udc_core: Checking for a failed platform_get_irq() call in fsl_udc_probe() https://lore.kernel.org/linux-usb/36341bb1-1e00-5eb1-d032-60dcc614ddaf@web.de/ https://lkml.org/lkml/2020/4/8/442 … > +++ b/drivers/usb/gadget/udc/fsl_udc_core.c > @@ -2441,8 +2441,8 @@ static int fsl_udc_probe(struct platform_device *pdev) > udc_controller->max_ep = (dccparams & DCCPARAMS_DEN_MASK) * 2; > > udc_controller->irq = platform_get_irq(pdev, 0); > - if (!udc_controller->irq) { > - ret = -ENODEV; > + if (udc_controller->irq <= 0) { Will such a failure predicate need any more clarification? How does this check fit to the current software documentation? > + ret = udc_controller->irq ? : -ENODEV; Will it be clearer to specify values for all cases in such a conditional operator (instead of leaving one case empty)? Regards, Markus
Hi Markus On 2020/4/10 15:33, Markus Elfring wrote: >> If the function "platform_get_irq()" failed, the negative value >> returned will not be detected here, including "-EPROBE_DEFER", > I suggest to adjust this change description. > > Wording alternative: > The negative return value (which could eventually be “-EPROBE_DEFER”) > will not be detected here from a failed call of the function “platform_get_irq”. Hardware experiments show that the negative return value is not just "-EPROBE_DEFER". > >> which causes the application to fail to get the correct error message. > Will another fine-tuning become relevant also for this wording? Maybe that's not quite accurate. > > >> Thus it must be fixed. > Wording alternative: > Thus adjust the error detection and corresponding exception handling. Got it. > > >> Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com> >> Signed-off-by: Shengju Zhang <zhangshengju@cmss.chinamobile.com> > How do you think about to add the tags “Fixes”, “Link” and “Reported-by”? > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=c0cc271173b2e1c2d8d0ceaef14e4dfa79eefc0d#n584 > > usb: gadget: fsl_udc_core: Checking for a failed platform_get_irq() call in fsl_udc_probe() > https://lore.kernel.org/linux-usb/36341bb1-1e00-5eb1-d032-60dcc614ddaf@web.de/ > https://lkml.org/lkml/2020/4/8/442 > > … >> +++ b/drivers/usb/gadget/udc/fsl_udc_core.c >> @@ -2441,8 +2441,8 @@ static int fsl_udc_probe(struct platform_device *pdev) >> udc_controller->max_ep = (dccparams & DCCPARAMS_DEN_MASK) * 2; >> >> udc_controller->irq = platform_get_irq(pdev, 0); >> - if (!udc_controller->irq) { >> - ret = -ENODEV; >> + if (udc_controller->irq <= 0) { > Will such a failure predicate need any more clarification? > > How does this check fit to the current software documentation? Maybe my tags are not suitable. > > >> + ret = udc_controller->irq ? : -ENODEV; > Will it be clearer to specify values for all cases in such a conditional operator > (instead of leaving one case empty)? I don't know what you mean of "instead of leaving one case empty". But by experiment, "ret = udc_controller->irq ? : -ENODEV" or "ret = udc_controller->irq < 0 ? udc_controller->irq : -ENODEV" should be suitable here. Thank you for your guidance. Tang Bin
> Hardware experiments show that the negative return value is not just "-EPROBE_DEFER". How much will this specific error code influence our understanding of the discussed software situation? >>> + ret = udc_controller->irq ? : -ENODEV; >> Will it be clearer to specify values for all cases in such a conditional operator >> (instead of leaving one case empty)? > > I don't know what you mean of "instead of leaving one case empty". I suggest to reconsider also the proposed specification “… ? : …”. Regards, Markus
Hi Markus: On 2020/4/10 16:30, Markus Elfring wrote: >> Hardware experiments show that the negative return value is not just "-EPROBE_DEFER". > How much will this specific error code influence our understanding > of the discussed software situation? > From my superficial knowledge, I think we should not think about it too complicated. The return value is just zero or negative, and for these negative return value, such as "-ENODEV"、"-ENXIO"、"-ENOENT"、“EPROBE_DEFER”，may be the same effect。But“-EPROBE_DEFER”has another importment function: Driver requested deferred probing，which is used in cases where the dependency resource is not ready during the driver initialization process. >>>> + ret = udc_controller->irq ? : -ENODEV; >>> Will it be clearer to specify values for all cases in such a conditional operator >>> (instead of leaving one case empty)? >> I don't know what you mean of "instead of leaving one case empty". > I suggest to reconsider also the proposed specification “… ? : …”. What you mean is the way I'm written？ I have provided two ways of patching, both functional can be verified on hardware. Thanks for your patience. Tang Bin
Hello! On 10.04.2020 4:58, Tang Bin wrote: > If the function "platform_get_irq()" failed, the negative value > returned will not be detected here, including "-EPROBE_DEFER", which > causes the application to fail to get the correct error message. > Thus it must be fixed. > > Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com> > Signed-off-by: Shengju Zhang <zhangshengju@cmss.chinamobile.com> > --- > drivers/usb/gadget/udc/fsl_udc_core.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/usb/gadget/udc/fsl_udc_core.c b/drivers/usb/gadget/udc/fsl_udc_core.c > index ec6eda426..60853ad10 100644 > --- a/drivers/usb/gadget/udc/fsl_udc_core.c > +++ b/drivers/usb/gadget/udc/fsl_udc_core.c > @@ -2441,8 +2441,8 @@ static int fsl_udc_probe(struct platform_device *pdev) > udc_controller->max_ep = (dccparams & DCCPARAMS_DEN_MASK) * 2; > > udc_controller->irq = platform_get_irq(pdev, 0); > - if (!udc_controller->irq) { This check has always been wrong, doesn't account for real errors and probe deferral... > - ret = -ENODEV; > + if (udc_controller->irq <= 0) { Note that platfrom_get_irq() no longer returns 0 on error. > + ret = udc_controller->irq ? : -ENODEV; > goto err_iounmap; > } MBR, Sergei
On 10.04.2020 4:58, Tang Bin wrote: > If the function "platform_get_irq()" failed, the negative value > returned will not be detected here, including "-EPROBE_DEFER", which > causes the application to fail to get the correct error message. > Thus it must be fixed. platform_get_irq() prints an appropriate error message, the problem is that the current code calls request_irq() with error code instead of IRQ. > Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com> > Signed-off-by: Shengju Zhang <zhangshengju@cmss.chinamobile.com> [...] MBR, Sergei
On Fri, Apr 10, 2020 at 04:05:06PM +0800, Tang Bin wrote: > > > > > > > Thus it must be fixed. > > Wording alternative: > > Thus adjust the error detection and corresponding exception handling. > > Got it. Wow... No, don't listen to Markus when it comes to writing commit messages. You couldn't find worse advice anywhere. :P regards, dan carpenter
Hi On 2020/4/14 16:30, Dan Carpenter wrote: > On Fri, Apr 10, 2020 at 04:05:06PM +0800, Tang Bin wrote: >>> >>>> Thus it must be fixed. >>> Wording alternative: >>> Thus adjust the error detection and corresponding exception handling. >> Got it. > Wow... > > No, don't listen to Markus when it comes to writing commit messages. > You couldn't find worse advice anywhere. :P I'm actively waiting for a reply from the maintainer. Thank you very much. Tang Bin
On Tue, Apr 14, 2020 at 4:13 AM Tang Bin <tangbin@cmss.chinamobile.com> wrote: > > Hi > > On 2020/4/14 16:30, Dan Carpenter wrote: > > On Fri, Apr 10, 2020 at 04:05:06PM +0800, Tang Bin wrote: > >>> > >>>> Thus it must be fixed. > >>> Wording alternative: > >>> Thus adjust the error detection and corresponding exception handling. > >> Got it. > > Wow... > > > > No, don't listen to Markus when it comes to writing commit messages. > > You couldn't find worse advice anywhere. :P > > I'm actively waiting for a reply from the maintainer. Thank you very much. Sorry for the late response. Acked-by: Li Yang <leoyang.li@nxp.com> Regards, Leo
diff --git a/drivers/usb/gadget/udc/fsl_udc_core.c b/drivers/usb/gadget/udc/fsl_udc_core.c index ec6eda426..60853ad10 100644 --- a/drivers/usb/gadget/udc/fsl_udc_core.c +++ b/drivers/usb/gadget/udc/fsl_udc_core.c @@ -2441,8 +2441,8 @@ static int fsl_udc_probe(struct platform_device *pdev) udc_controller->max_ep = (dccparams & DCCPARAMS_DEN_MASK) * 2; udc_controller->irq = platform_get_irq(pdev, 0); - if (!udc_controller->irq) { - ret = -ENODEV; + if (udc_controller->irq <= 0) { + ret = udc_controller->irq ? : -ENODEV; goto err_iounmap; }