| Message ID | 14300712.5chtEzsjIN@wasted.cogentembedded.com |
|---|---|
| State | New |
| Headers | show |
On Sat, Mar 5, 2016 at 5:58 AM, Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> wrote: > The 'ravb' Ethernet driver couldn't connect to the PHY as the MDIO bus > appeared empty on the Renesas R-Car boards. The bug hunt finally pointed > at the commit adding the "init" pintcrl state: it tries to switch to non- > default state before the driver probe which should fail but doesn't as the > PFC pinctrl driver happens to call pinctrl_provide_dummies() which makes > all state lookups succeed, even though the state doesn't really exist. > That feature is only relevant to non-DT systems and all the ARM boards > that use the PFC driver have been converted to the DT boot, so limiting > it to the SuperH architecture seems The Right Thing... > > Fixes: ef0eebc05130 ("drivers/pinctrl: Add the concept of an "init" state") > Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> > CC: stable@vger.kernel.org Putting this on hold because another patch is being discussed as a more proper fix. Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hello. On 3/9/2016 2:09 PM, Linus Walleij wrote: >> The 'ravb' Ethernet driver couldn't connect to the PHY as the MDIO bus >> appeared empty on the Renesas R-Car boards. The bug hunt finally pointed >> at the commit adding the "init" pintcrl state: it tries to switch to non- >> default state before the driver probe which should fail but doesn't as the >> PFC pinctrl driver happens to call pinctrl_provide_dummies() which makes >> all state lookups succeed, even though the state doesn't really exist. >> That feature is only relevant to non-DT systems and all the ARM boards >> that use the PFC driver have been converted to the DT boot, so limiting >> it to the SuperH architecture seems The Right Thing... >> >> Fixes: ef0eebc05130 ("drivers/pinctrl: Add the concept of an "init" state") >> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> >> CC: stable@vger.kernel.org > > Putting this on hold because another patch is being discussed as a > more proper fix. Sigh, I had that variant of the fix too but erroneously thought that of_have_populated_dt() could only be used with CONFIG_OF=y, so didn't publish that variant... Had no idea someone else's working on the same issue. > Yours, > Linus Walleij MBR, Sergei -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Index: linux-pinctrl/drivers/pinctrl/sh-pfc/core.c =================================================================== --- linux-pinctrl.orig/drivers/pinctrl/sh-pfc/core.c +++ linux-pinctrl/drivers/pinctrl/sh-pfc/core.c @@ -545,7 +545,8 @@ static int sh_pfc_probe(struct platform_ return ret; } - pinctrl_provide_dummies(); + if (IS_ENABLED(CONFIG_SUPERH)) + pinctrl_provide_dummies(); ret = sh_pfc_init_ranges(pfc); if (ret < 0)
The 'ravb' Ethernet driver couldn't connect to the PHY as the MDIO bus appeared empty on the Renesas R-Car boards. The bug hunt finally pointed at the commit adding the "init" pintcrl state: it tries to switch to non- default state before the driver probe which should fail but doesn't as the PFC pinctrl driver happens to call pinctrl_provide_dummies() which makes all state lookups succeed, even though the state doesn't really exist. That feature is only relevant to non-DT systems and all the ARM boards that use the PFC driver have been converted to the DT boot, so limiting it to the SuperH architecture seems The Right Thing... Fixes: ef0eebc05130 ("drivers/pinctrl: Add the concept of an "init" state") Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> CC: stable@vger.kernel.org --- The patch is against the 'fixes' branch of Linus W.'s 'linux-pinctrl.git' repo. As the bug only seems to happen to the hardware that's not pre-configured by U-Boot, this patch can wait till 4.6 in principle... drivers/pinctrl/sh-pfc/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html