| Message ID | 20191001180834.1158-1-geert+renesas@glider.be (mailing list archive) |
|---|---|
| State | Mainlined |
| Commit | faf97b84fa86d40eba5a30e9071dfb55133fb1b7 |
| Delegated to: | Geert Uytterhoeven |
| Headers | show |
| Series | mmc: sh_mmcif: Use platform_get_irq_optional() for optional interrupt | expand |
Hi Geert-san, > From: Geert Uytterhoeven, Sent: Wednesday, October 2, 2019 3:09 AM > > As platform_get_irq() now prints an error when the interrupt does not > exist, a scary warning may be printed for an optional interrupt: > > sh_mmcif ee200000.mmc: IRQ index 1 not found > > Fix this by calling platform_get_irq_optional() instead for the second > interrupt, which is optional. > > Remove the now superfluous error printing for the first interrupt, which > is mandatory. > > Fixes: 7723f4c5ecdb8d83 ("driver core: platform: Add an error message to platform_get_irq*()") > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> > --- > This is a fix for v5.4-rc1. > --- Thank you for the patch! Reviewed-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> And, I tested this patch on R-Car H2. So, Tested-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Best regards, Yoshihiro Shimoda
On Tue, Oct 01, 2019 at 08:08:34PM +0200, Geert Uytterhoeven wrote: > As platform_get_irq() now prints an error when the interrupt does not > exist, a scary warning may be printed for an optional interrupt: > > sh_mmcif ee200000.mmc: IRQ index 1 not found > > Fix this by calling platform_get_irq_optional() instead for the second > interrupt, which is optional. > > Remove the now superfluous error printing for the first interrupt, which > is mandatory. > > Fixes: 7723f4c5ecdb8d83 ("driver core: platform: Add an error message to platform_get_irq*()") > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Tested on a R-Car H2: it does make the error message go away and the MMCIF device acts normal during boot. Can't enter userspace currently with v5.4-rc1 but this is unrelated to this patch and MMCIF. Looks like a configuration thing on my side, so I will still give: Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com> But will report back when I fixed the unrelated issue.
Hi, > From: Wolfram Sang, Sent: Wednesday, October 2, 2019 5:16 PM <snip> > Tested on a R-Car H2: it does make the error message go away and the > MMCIF device acts normal during boot. Can't enter userspace currently > with v5.4-rc1 but this is unrelated to this patch and MMCIF. Looks like a > configuration thing on my side, so I will still give: I also should have reported this though, my environment (R-Car H2 + NFS + buildroot on v5.4-rc1 with shmobile_defconfig) also has a similar issue like the following: [ 3.573488] VFS: Mounted root (nfs filesystem) on device 0:16. [ 3.579869] devtmpfs: mounted [ 3.588014] Freeing unused kernel memory: 1024K [ 3.651771] Run /sbin/init as init process Starting syslogd: OK Starting klogd: OK Initializing random number generator... [ 4.073629] random: dd: uninitialized urandom read (512 bytes read) urandom start: failed. done. Starting network: ip: OVERRUN: Bad address ip: OVERRUN: Bad address ip: OVERRUN: Bad address Best regards, Yoshihiro Shimoda
Hi Shimoda-san, Wolfram, On Wed, Oct 2, 2019 at 10:26 AM Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> wrote: > > From: Wolfram Sang, Sent: Wednesday, October 2, 2019 5:16 PM > <snip> > > Tested on a R-Car H2: it does make the error message go away and the > > MMCIF device acts normal during boot. Can't enter userspace currently > > with v5.4-rc1 but this is unrelated to this patch and MMCIF. Looks like a > > configuration thing on my side, so I will still give: > > I also should have reported this though, my environment (R-Car H2 + NFS + > buildroot on v5.4-rc1 with shmobile_defconfig) also has a similar issue > like the following: > > [ 3.573488] VFS: Mounted root (nfs filesystem) on device 0:16. > [ 3.579869] devtmpfs: mounted > [ 3.588014] Freeing unused kernel memory: 1024K > [ 3.651771] Run /sbin/init as init process > Starting syslogd: OK > Starting klogd: OK > Initializing random number generator... [ 4.073629] random: dd: uninitialized urandom read (512 bytes read) > urandom start: failed. > done. > Starting network: ip: OVERRUN: Bad address > ip: OVERRUN: Bad address > ip: OVERRUN: Bad address Please cherry-pick the top commit from renesas-devel: 6e47c841329eb9b0 ("ARM: fix __get_user_check() in case uaccess_* calls are not inlined") Gr{oetje,eeting}s, Geert
> Please cherry-pick the top commit from renesas-devel: > 6e47c841329eb9b0 ("ARM: fix __get_user_check() in case uaccess_* calls > are not inlined") Thanks a lot, Geert. I can confirm this patch fixes the issue I was seeing: Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Could you kindly forward this tag to Yamada-san or bounce me the patch so I can do it? Thanks again, Wolfram
Hi Wolfram, On Wed, Oct 2, 2019 at 10:58 AM Wolfram Sang <wsa@the-dreams.de> wrote: > > Please cherry-pick the top commit from renesas-devel: > > 6e47c841329eb9b0 ("ARM: fix __get_user_check() in case uaccess_* calls > > are not inlined") > > Thanks a lot, Geert. I can confirm this patch fixes the issue I was > seeing: > > Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com> > > Could you kindly forward this tag to Yamada-san or bounce me the patch > so I can do it? Please check your linux-renesas-soc folder. I CCed the list when providing my own Tested-by yesterday. Gr{oetje,eeting}s, Geert
Hi Geert-san, > From: Geert Uytterhoeven, Sent: Wednesday, October 2, 2019 5:32 PM > > Hi Shimoda-san, Wolfram, > > On Wed, Oct 2, 2019 at 10:26 AM Yoshihiro Shimoda > <yoshihiro.shimoda.uh@renesas.com> wrote: > > > From: Wolfram Sang, Sent: Wednesday, October 2, 2019 5:16 PM > > <snip> > > > Tested on a R-Car H2: it does make the error message go away and the > > > MMCIF device acts normal during boot. Can't enter userspace currently > > > with v5.4-rc1 but this is unrelated to this patch and MMCIF. Looks like a > > > configuration thing on my side, so I will still give: > > > > I also should have reported this though, my environment (R-Car H2 + NFS + > > buildroot on v5.4-rc1 with shmobile_defconfig) also has a similar issue > > like the following: > > > > [ 3.573488] VFS: Mounted root (nfs filesystem) on device 0:16. > > [ 3.579869] devtmpfs: mounted > > [ 3.588014] Freeing unused kernel memory: 1024K > > [ 3.651771] Run /sbin/init as init process > > Starting syslogd: OK > > Starting klogd: OK > > Initializing random number generator... [ 4.073629] random: dd: uninitialized urandom read (512 bytes read) > > urandom start: failed. > > done. > > Starting network: ip: OVERRUN: Bad address > > ip: OVERRUN: Bad address > > ip: OVERRUN: Bad address > > Please cherry-pick the top commit from renesas-devel: > 6e47c841329eb9b0 ("ARM: fix __get_user_check() in case uaccess_* calls > are not inlined") Thank you for the information. This commit could fix the issue on my environment. Best regards, Yoshihiro Shimoda
On Tue, 1 Oct 2019 at 20:08, Geert Uytterhoeven <geert+renesas@glider.be> wrote: > > As platform_get_irq() now prints an error when the interrupt does not > exist, a scary warning may be printed for an optional interrupt: > > sh_mmcif ee200000.mmc: IRQ index 1 not found > > Fix this by calling platform_get_irq_optional() instead for the second > interrupt, which is optional. > > Remove the now superfluous error printing for the first interrupt, which > is mandatory. > > Fixes: 7723f4c5ecdb8d83 ("driver core: platform: Add an error message to platform_get_irq*()") > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Applied for fixes, thanks! Kind regards Uffe > --- > This is a fix for v5.4-rc1. > --- > drivers/mmc/host/sh_mmcif.c | 6 ++---- > 1 file changed, 2 insertions(+), 4 deletions(-) > > diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c > index 81bd9afb0980525e..98c575de43c755ed 100644 > --- a/drivers/mmc/host/sh_mmcif.c > +++ b/drivers/mmc/host/sh_mmcif.c > @@ -1393,11 +1393,9 @@ static int sh_mmcif_probe(struct platform_device *pdev) > const char *name; > > irq[0] = platform_get_irq(pdev, 0); > - irq[1] = platform_get_irq(pdev, 1); > - if (irq[0] < 0) { > - dev_err(dev, "Get irq error\n"); > + irq[1] = platform_get_irq_optional(pdev, 1); > + if (irq[0] < 0) > return -ENXIO; > - } > > res = platform_get_resource(pdev, IORESOURCE_MEM, 0); > reg = devm_ioremap_resource(dev, res); > -- > 2.17.1 >
diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c index 81bd9afb0980525e..98c575de43c755ed 100644 --- a/drivers/mmc/host/sh_mmcif.c +++ b/drivers/mmc/host/sh_mmcif.c @@ -1393,11 +1393,9 @@ static int sh_mmcif_probe(struct platform_device *pdev) const char *name; irq[0] = platform_get_irq(pdev, 0); - irq[1] = platform_get_irq(pdev, 1); - if (irq[0] < 0) { - dev_err(dev, "Get irq error\n"); + irq[1] = platform_get_irq_optional(pdev, 1); + if (irq[0] < 0) return -ENXIO; - } res = platform_get_resource(pdev, IORESOURCE_MEM, 0); reg = devm_ioremap_resource(dev, res);
As platform_get_irq() now prints an error when the interrupt does not exist, a scary warning may be printed for an optional interrupt: sh_mmcif ee200000.mmc: IRQ index 1 not found Fix this by calling platform_get_irq_optional() instead for the second interrupt, which is optional. Remove the now superfluous error printing for the first interrupt, which is mandatory. Fixes: 7723f4c5ecdb8d83 ("driver core: platform: Add an error message to platform_get_irq*()") Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> --- This is a fix for v5.4-rc1. --- drivers/mmc/host/sh_mmcif.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)