| Message ID | 20220721172935.13864-1-j@jannau.net |
|---|---|
| State | Superseded |
| Delegated to: | Marek Vasut |
| Headers | show |
| Series | [1/1] usb: storage: stop probe on "Invalid device" | expand |
Hi Marek, On Thu, 21 Jul 2022 at 11:29, Janne Grunau <j@jannau.net> wrote: > > Fixes a crash during probing of sd card readers without medium present. > > Link: https://github.com/AsahiLinux/linux/issues/44 > Signed-off-by: Janne Grunau <j@jannau.net> > --- > common/usb_storage.c | 4 +--- > 1 file changed, 1 insertion(+), 3 deletions(-) Reviewed-by: Simon Glass <sjg@chromium.org> Shall I pick this one up? Regards, Simon
Hi Simon, On Sun, Aug 07, 2022 at 09:47:56AM -0600, Simon Glass wrote: > Hi Marek, > > On Thu, 21 Jul 2022 at 11:29, Janne Grunau <j@jannau.net> wrote: > > > > Fixes a crash during probing of sd card readers without medium present. > > > > Link: https://github.com/AsahiLinux/linux/issues/44 > > Signed-off-by: Janne Grunau <j@jannau.net> > > --- > > common/usb_storage.c | 4 +--- > > 1 file changed, 1 insertion(+), 3 deletions(-) > > Reviewed-by: Simon Glass <sjg@chromium.org> I made this comment on another bug report: https://lists.denx.de/pipermail/u-boot/2022-July/489717.html I think we should continue to scan the *bus* even if detecting/activating a specific device fails. Moreover, we would better do full recovery from blk_create_device(f)() as it has some side-effects, including an increased "devnum". -Takahiro Akashi > Shall I pick this one up? > > Regards, > Simon
Hi Takahiro, On Sun, 7 Aug 2022 at 20:18, AKASHI Takahiro <takahiro.akashi@linaro.org> wrote: > > Hi Simon, > > On Sun, Aug 07, 2022 at 09:47:56AM -0600, Simon Glass wrote: > > Hi Marek, > > > > On Thu, 21 Jul 2022 at 11:29, Janne Grunau <j@jannau.net> wrote: > > > > > > Fixes a crash during probing of sd card readers without medium present. > > > > > > Link: https://github.com/AsahiLinux/linux/issues/44 > > > Signed-off-by: Janne Grunau <j@jannau.net> > > > --- > > > common/usb_storage.c | 4 +--- > > > 1 file changed, 1 insertion(+), 3 deletions(-) > > > > Reviewed-by: Simon Glass <sjg@chromium.org> > > I made this comment on another bug report: > https://lists.denx.de/pipermail/u-boot/2022-July/489717.html > > I think we should continue to scan the *bus* even if detecting/activating > a specific device fails. > > Moreover, we would better do full recovery from blk_create_device(f)() > as it has some side-effects, including an increased "devnum". OK I didn't see that. Can you please send an alternative patch? > > -Takahiro Akashi > > > Shall I pick this one up? Regards, Simon
diff --git a/common/usb_storage.c b/common/usb_storage.c index eaa31374ef73..5a62d61f4382 100644 --- a/common/usb_storage.c +++ b/common/usb_storage.c @@ -236,9 +236,7 @@ static int usb_stor_probe_device(struct usb_device *udev) debug("%s: Found device %p\n", __func__, udev); } else { debug("usb_stor_get_info: Invalid device\n"); - ret = device_unbind(dev); - if (ret) - return ret; + return device_unbind(dev); } ret = blk_probe_or_unbind(dev);
Fixes a crash during probing of sd card readers without medium present. Link: https://github.com/AsahiLinux/linux/issues/44 Signed-off-by: Janne Grunau <j@jannau.net> --- common/usb_storage.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)