| Message ID | 20190516090657.25211-2-j-keerthy@ti.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | None | expand |
Hi, * Keerthy <j-keerthy@ti.com> [190516 09:06]: > rtc is fused out on dra76 and accessing target module > register is causing a boot crash hence disable it. So for a fix, can we have a separate dra7 something dtsi file to disable these instead? Or are there already multiple SoC revisions for the same EVM? Then in the long run, if there are the same EVMs with multiple SoC options, the best thing to do is to would be to detect the SoC type and update the property dynamically to set the features not available on the booted SoC to status = "disabled". Seems like that could be done in the ti-sysc driver probe unless needed earlier. Regards, Tony
On 5/16/2019 9:42 PM, Tony Lindgren wrote: > Hi, > > * Keerthy <j-keerthy@ti.com> [190516 09:06]: >> rtc is fused out on dra76 and accessing target module >> register is causing a boot crash hence disable it. > > So for a fix, can we have a separate dra7 something dtsi file > to disable these instead? > > Or are there already multiple SoC revisions for the same EVM? dra76 & dra71 have rtc fused out. So i did not introduce a new dtsi file to disable. > > Then in the long run, if there are the same EVMs with multiple > SoC options, the best thing to do is to would be to detect the > SoC type and update the property dynamically to set the features > not available on the booted SoC to status = "disabled". Seems > like that could be done in the ti-sysc driver probe unless needed > earlier. For now rtc is disabled only in dra71/dra76. So best disable it in the evm.dts? Not sure if we need dynamic disabling as we know at DT level that it is to be disabled. > > Regards, > > Tony >
* keerthy <j-keerthy@ti.com> [190516 16:31]: > > > On 5/16/2019 9:42 PM, Tony Lindgren wrote: > > Hi, > > > > * Keerthy <j-keerthy@ti.com> [190516 09:06]: > > > rtc is fused out on dra76 and accessing target module > > > register is causing a boot crash hence disable it. > > > > So for a fix, can we have a separate dra7 something dtsi file > > to disable these instead? > > > > Or are there already multiple SoC revisions for the same EVM? > > dra76 & dra71 have rtc fused out. So i did not introduce a new dtsi file > to disable. But then any new board with dra76 or dra71 will need to debug the same issue again. Sure we can get away for now tweaking the board file, but to me it sounds like it's going to be more devices that will be affected too? Is there some feature matrix available somewhere online? > > Then in the long run, if there are the same EVMs with multiple > > SoC options, the best thing to do is to would be to detect the > > SoC type and update the property dynamically to set the features > > not available on the booted SoC to status = "disabled". Seems > > like that could be done in the ti-sysc driver probe unless needed > > earlier. > > For now rtc is disabled only in dra71/dra76. So best disable it in the > evm.dts? Not sure if we need dynamic disabling as we know at DT level that > it is to be disabled. Well the thing is we should make introducing new board dts files as easy as including the SoC dtsi file and with that it should boot with no extra debugging. How about add minimal dra76 and dra71 dtsi files in addition to the board specific fix(es)? Then for v5.3, we can deal adding more dra7 specifc evm files using these dtsi files or dynamically start disabling modules. Regards, Tony
On 5/16/2019 10:17 PM, Tony Lindgren wrote: > * keerthy <j-keerthy@ti.com> [190516 16:31]: >> >> >> On 5/16/2019 9:42 PM, Tony Lindgren wrote: >>> Hi, >>> >>> * Keerthy <j-keerthy@ti.com> [190516 09:06]: >>>> rtc is fused out on dra76 and accessing target module >>>> register is causing a boot crash hence disable it. >>> >>> So for a fix, can we have a separate dra7 something dtsi file >>> to disable these instead? >>> >>> Or are there already multiple SoC revisions for the same EVM? >> >> dra76 & dra71 have rtc fused out. So i did not introduce a new dtsi file >> to disable. > > But then any new board with dra76 or dra71 will need to debug > the same issue again. Sure we can get away for now tweaking the > board file, but to me it sounds like it's going to be more > devices that will be affected too? Okay. This is a SoC related issue so yes any new board will have to again implement disabling. > > Is there some feature matrix available somewhere online? Not that i know of. I will try finding something. > >>> Then in the long run, if there are the same EVMs with multiple >>> SoC options, the best thing to do is to would be to detect the >>> SoC type and update the property dynamically to set the features >>> not available on the booted SoC to status = "disabled". Seems >>> like that could be done in the ti-sysc driver probe unless needed >>> earlier. >> >> For now rtc is disabled only in dra71/dra76. So best disable it in the >> evm.dts? Not sure if we need dynamic disabling as we know at DT level that >> it is to be disabled. > > Well the thing is we should make introducing new board dts files > as easy as including the SoC dtsi file and with that it should > boot with no extra debugging. > > How about add minimal dra76 and dra71 dtsi files in addition to > the board specific fix(es)? Then for v5.3, we can deal adding more > dra7 specifc evm files using these dtsi files or dynamically start > disabling modules. You mean having a dra76/dra71.dtsi with rtc/usb4_tm disabled and that gets included in dra76/dra71-evm.dts? > > Regards, > > Tony >
* keerthy <j-keerthy@ti.com> [190516 16:54]: > On 5/16/2019 10:17 PM, Tony Lindgren wrote: > > But then any new board with dra76 or dra71 will need to debug > > the same issue again. Sure we can get away for now tweaking the > > board file, but to me it sounds like it's going to be more > > devices that will be affected too? > > Okay. This is a SoC related issue so yes any new board will have to again > implement disabling. OK > > Is there some feature matrix available somewhere online? > > Not that i know of. I will try finding something. OK thanks. > > How about add minimal dra76 and dra71 dtsi files in addition to > > the board specific fix(es)? Then for v5.3, we can deal adding more > > dra7 specifc evm files using these dtsi files or dynamically start > > disabling modules. > > You mean having a dra76/dra71.dtsi with rtc/usb4_tm disabled and that gets > included in dra76/dra71-evm.dts? Yeah, add dra76.dtsi and dra71.dtsi, then include those. If however you have the same evm dts file with multiple SoC variants available, then for now you need to also disable devices in the evm dts. Or add another evm dts file. Regards, Tony
diff --git a/arch/arm/boot/dts/dra7-l4.dtsi b/arch/arm/boot/dts/dra7-l4.dtsi index fe9f0bc29fec..3b4cba9da91f 100644 --- a/arch/arm/boot/dts/dra7-l4.dtsi +++ b/arch/arm/boot/dts/dra7-l4.dtsi @@ -3543,7 +3543,7 @@ }; }; - target-module@38000 { /* 0x48838000, ap 29 12.0 */ + rtctarget: target-module@38000 { /* 0x48838000, ap 29 12.0 */ compatible = "ti,sysc-omap4-simple", "ti,sysc"; ti,hwmods = "rtcss"; reg = <0x38074 0x4>, diff --git a/arch/arm/boot/dts/dra76-evm.dts b/arch/arm/boot/dts/dra76-evm.dts index 6607fbe3c67c..644fe1961d28 100644 --- a/arch/arm/boot/dts/dra76-evm.dts +++ b/arch/arm/boot/dts/dra76-evm.dts @@ -363,6 +363,10 @@ pinctrl-3 = <&mmc4_pins_hs &mmc4_iodelay_manual1_conf>; }; +&rtctarget { + status = "disabled"; +}; + /* No RTC on this device */ &rtc { status = "disabled";
rtc is fused out on dra76 and accessing target module register is causing a boot crash hence disable it. Fixes: 549fce068a3112 ("ARM: dts: dra7: Add l4 interconnect hierarchy and ti-sysc data") Signed-off-by: Keerthy <j-keerthy@ti.com> --- arch/arm/boot/dts/dra7-l4.dtsi | 2 +- arch/arm/boot/dts/dra76-evm.dts | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-)