| Message ID | 1505230349-5412-1-git-send-email-philipp.tomsich@theobroma-systems.com |
|---|---|
| Headers | show |
| Series | rockchip: enable live tree on the RK3399 and validate on the RK3399-Q7 | expand |
On 09/12/2017 05:32 PM, Philipp Tomsich wrote: > This series repeats the same exercise as for the RK3368 and adapts any > lingering dependencies necessary to move the RK3399-based boards over > to use a live tree. Validated on the RK3399-Q7 (Puma). > > Resent, due to an apparent patchworks hick-up yesterday that "ate my series". > > > Philipp Tomsich (6): > rockchip: clk: rk3399: Convert to livetree > rockchip: xhci: Convert to livetree > rockchip: efuse: change to use dev_read_addr_ptr > usb: dwc2: convert to livetree > usb: host: ehci-generic: convert to livetree > rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) So how are these patches dependent on one another ? They look like mostly separate patches to me. I certainly cannot take the whole patchset through -usb ...
> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: > > On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >> This series repeats the same exercise as for the RK3368 and adapts any >> lingering dependencies necessary to move the RK3399-based boards over >> to use a live tree. Validated on the RK3399-Q7 (Puma). >> >> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >> >> >> Philipp Tomsich (6): >> rockchip: clk: rk3399: Convert to livetree >> rockchip: xhci: Convert to livetree >> rockchip: efuse: change to use dev_read_addr_ptr >> usb: dwc2: convert to livetree >> usb: host: ehci-generic: convert to livetree >> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) > > So how are these patches dependent on one another ? They look like > mostly separate patches to me. I certainly cannot take the whole > patchset through -usb ... These are all dependencies to move RK3399 boards to OF_LIVE. We have two options: (a) keep them together and I move them through u-boot-rockchip after you send a review for the usb-related patches (b) I split the series, but then the usb-related changes are left dangling as there won’t be a clear indication how this was tested and what context these came from. Either approach is fine with me. Regards, Philipp.
> On 13 Sep 2017, at 11:47, Dr. Philipp Tomsich <philipp.tomsich@theobroma-systems.com> wrote: > > >> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: >> >> On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >>> This series repeats the same exercise as for the RK3368 and adapts any >>> lingering dependencies necessary to move the RK3399-based boards over >>> to use a live tree. Validated on the RK3399-Q7 (Puma). >>> >>> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >>> >>> >>> Philipp Tomsich (6): >>> rockchip: clk: rk3399: Convert to livetree >>> rockchip: xhci: Convert to livetree >>> rockchip: efuse: change to use dev_read_addr_ptr >>> usb: dwc2: convert to livetree >>> usb: host: ehci-generic: convert to livetree >>> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) >> >> So how are these patches dependent on one another ? They look like >> mostly separate patches to me. I certainly cannot take the whole >> patchset through -usb ... > > These are all dependencies to move RK3399 boards to OF_LIVE. > > We have two options: > (a) keep them together and I move them through u-boot-rockchip after > you send a review for the usb-related patches > (b) I split the series, but then the usb-related changes are left dangling > as there won’t be a clear indication how this was tested and what > context these came from. > > Either approach is fine with me. I just realised that the entire series has been delegated to me in Patchwork anyway, so it seems as if option (a) would be the “default choice” … Regards, Philipp.
On 09/13/2017 11:47 AM, Dr. Philipp Tomsich wrote: > >> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: >> >> On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >>> This series repeats the same exercise as for the RK3368 and adapts any >>> lingering dependencies necessary to move the RK3399-based boards over >>> to use a live tree. Validated on the RK3399-Q7 (Puma). >>> >>> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >>> >>> >>> Philipp Tomsich (6): >>> rockchip: clk: rk3399: Convert to livetree >>> rockchip: xhci: Convert to livetree >>> rockchip: efuse: change to use dev_read_addr_ptr >>> usb: dwc2: convert to livetree >>> usb: host: ehci-generic: convert to livetree >>> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) >> >> So how are these patches dependent on one another ? They look like >> mostly separate patches to me. I certainly cannot take the whole >> patchset through -usb ... > > These are all dependencies to move RK3399 boards to OF_LIVE. > > We have two options: > (a) keep them together and I move them through u-boot-rockchip after > you send a review for the usb-related patches > (b) I split the series, but then the usb-related changes are left dangling > as there won’t be a clear indication how this was tested and what > context these came from. > > Either approach is fine with me. Since some of this is core USB stuff, it should go through -usb , that's what the subsystem trees are for. I also don't see any problem with pulling the USB ones in separately, they do not seem to depend on anything else in that series.
On 09/13/2017 11:48 AM, Dr. Philipp Tomsich wrote: > >> On 13 Sep 2017, at 11:47, Dr. Philipp Tomsich <philipp.tomsich@theobroma-systems.com> wrote: >> >> >>> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: >>> >>> On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >>>> This series repeats the same exercise as for the RK3368 and adapts any >>>> lingering dependencies necessary to move the RK3399-based boards over >>>> to use a live tree. Validated on the RK3399-Q7 (Puma). >>>> >>>> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >>>> >>>> >>>> Philipp Tomsich (6): >>>> rockchip: clk: rk3399: Convert to livetree >>>> rockchip: xhci: Convert to livetree >>>> rockchip: efuse: change to use dev_read_addr_ptr >>>> usb: dwc2: convert to livetree >>>> usb: host: ehci-generic: convert to livetree >>>> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) >>> >>> So how are these patches dependent on one another ? They look like >>> mostly separate patches to me. I certainly cannot take the whole >>> patchset through -usb ... >> >> These are all dependencies to move RK3399 boards to OF_LIVE. >> >> We have two options: >> (a) keep them together and I move them through u-boot-rockchip after >> you send a review for the usb-related patches >> (b) I split the series, but then the usb-related changes are left dangling >> as there won’t be a clear indication how this was tested and what >> context these came from. >> >> Either approach is fine with me. > > I just realised that the entire series has been delegated to me in Patchwork > anyway, so it seems as if option (a) would be the “default choice” … And if SoC trees decide to pull in subsystem stuff, then subsystem maintainers will go nuts from dealing with changes in their area which is out of their control, there will be merge conflicts all over the place etc .. excellent.
> On 13 Sep 2017, at 11:49, Marek Vasut <marex@denx.de> wrote: > > On 09/13/2017 11:47 AM, Dr. Philipp Tomsich wrote: >> >>> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: >>> >>> On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >>>> This series repeats the same exercise as for the RK3368 and adapts any >>>> lingering dependencies necessary to move the RK3399-based boards over >>>> to use a live tree. Validated on the RK3399-Q7 (Puma). >>>> >>>> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >>>> >>>> >>>> Philipp Tomsich (6): >>>> rockchip: clk: rk3399: Convert to livetree >>>> rockchip: xhci: Convert to livetree >>>> rockchip: efuse: change to use dev_read_addr_ptr >>>> usb: dwc2: convert to livetree >>>> usb: host: ehci-generic: convert to livetree >>>> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) >>> >>> So how are these patches dependent on one another ? They look like >>> mostly separate patches to me. I certainly cannot take the whole >>> patchset through -usb ... >> >> These are all dependencies to move RK3399 boards to OF_LIVE. >> >> We have two options: >> (a) keep them together and I move them through u-boot-rockchip after >> you send a review for the usb-related patches >> (b) I split the series, but then the usb-related changes are left dangling >> as there won’t be a clear indication how this was tested and what >> context these came from. >> >> Either approach is fine with me. > > Since some of this is core USB stuff, it should go through -usb , that's > what the subsystem trees are for. I also don't see any problem with > pulling the USB ones in separately, they do not seem to depend on > anything else in that series. Just ACK those that you’ll take and I’ll try not to step on your toes when handling the remainder of the series. Regards, Philipp.
On 09/13/2017 11:53 AM, Dr. Philipp Tomsich wrote: > >> On 13 Sep 2017, at 11:49, Marek Vasut <marex@denx.de> wrote: >> >> On 09/13/2017 11:47 AM, Dr. Philipp Tomsich wrote: >>> >>>> On 13 Sep 2017, at 11:42, Marek Vasut <marex@denx.de> wrote: >>>> >>>> On 09/12/2017 05:32 PM, Philipp Tomsich wrote: >>>>> This series repeats the same exercise as for the RK3368 and adapts any >>>>> lingering dependencies necessary to move the RK3399-based boards over >>>>> to use a live tree. Validated on the RK3399-Q7 (Puma). >>>>> >>>>> Resent, due to an apparent patchworks hick-up yesterday that "ate my series". >>>>> >>>>> >>>>> Philipp Tomsich (6): >>>>> rockchip: clk: rk3399: Convert to livetree >>>>> rockchip: xhci: Convert to livetree >>>>> rockchip: efuse: change to use dev_read_addr_ptr >>>>> usb: dwc2: convert to livetree >>>>> usb: host: ehci-generic: convert to livetree >>>>> rockchip: defconfig: puma-rk3399: enable OF_LIVE (live tree) >>>> >>>> So how are these patches dependent on one another ? They look like >>>> mostly separate patches to me. I certainly cannot take the whole >>>> patchset through -usb ... >>> >>> These are all dependencies to move RK3399 boards to OF_LIVE. >>> >>> We have two options: >>> (a) keep them together and I move them through u-boot-rockchip after >>> you send a review for the usb-related patches >>> (b) I split the series, but then the usb-related changes are left dangling >>> as there won’t be a clear indication how this was tested and what >>> context these came from. >>> >>> Either approach is fine with me. >> >> Since some of this is core USB stuff, it should go through -usb , that's >> what the subsystem trees are for. I also don't see any problem with >> pulling the USB ones in separately, they do not seem to depend on >> anything else in that series. > > Just ACK those that you’ll take and I’ll try not to step on your toes when > handling the remainder of the series. I applied 2,4,5 to -usb , next time try to split the series please.