| Message ID | 434ba2467dd0cd011565625aeb3450650afe0aae.1611904394.git.viresh.kumar@linaro.org |
|---|---|
| State | Accepted |
| Commit | ce88c9c79455fb69c6843b4bd30196fccb978a57 |
| Headers | show |
| Series | dt: build overlays | expand |
Hi Viresh, Thanks for your patch (which I only noticed because it appeared in dt-rh/for-next ;-) On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > Add support for building DT overlays (%.dtbo). The overlay's source file > will have the usual extension, i.e. .dts, though the blob will have Why use .dts and not .dtso for overlays? Because you originally (until v5) had a single rule for building .dtb and .dtbo files? > .dtbo extension to distinguish it from normal blobs. > > Acked-by: Masahiro Yamada <masahiroy@kernel.org> > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org> Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On 05-02-21, 10:02, Geert Uytterhoeven wrote: > Hi Viresh, > > Thanks for your patch > (which I only noticed because it appeared in dt-rh/for-next ;-) > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > Add support for building DT overlays (%.dtbo). The overlay's source file > > will have the usual extension, i.e. .dts, though the blob will have > > Why use .dts and not .dtso for overlays? > Because you originally (until v5) had a single rule for building .dtb > and .dtbo files? I am fine with doing that as well if Rob and David agree to it. Rob did suggest that at one point but we didn't do much about it later on for some reason. FWIW, this will also require a change in the DTC compiler. -- viresh
Hi Viresh, On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > Thanks for your patch > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > will have the usual extension, i.e. .dts, though the blob will have > > > > Why use .dts and not .dtso for overlays? > > Because you originally (until v5) had a single rule for building .dtb > > and .dtbo files? > > I am fine with doing that as well if Rob and David agree to it. Rob > did suggest that at one point but we didn't do much about it later on > for some reason. > > FWIW, this will also require a change in the DTC compiler. Care to explain why? I've been using .dtsi for ages in https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On 05-02-21, 10:41, Geert Uytterhoeven wrote: > Hi Viresh, > > On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > > Thanks for your patch > > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > > will have the usual extension, i.e. .dts, though the blob will have > > > > > > Why use .dts and not .dtso for overlays? > > > Because you originally (until v5) had a single rule for building .dtb > > > and .dtbo files? > > > > I am fine with doing that as well if Rob and David agree to it. Rob > > did suggest that at one point but we didn't do much about it later on > > for some reason. > > > > FWIW, this will also require a change in the DTC compiler. > > Care to explain why? I've been using .dtsi for ages in > https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays I don't see you building them anywhere, they aren't added to the Makefile ever. What am I missing ? actually none of the dtso's were added to any makefile in that branch. Anyway, the DTC needs to know how to treat the dtso format and it will error out currently with unknown format kind of errors. Below email [1] have some information on the kind of changes required here. Also note that we had to do similar changes for dtbo earlier [2]. -- viresh [1] https://lore.kernel.org/lkml/CAK7LNASViCOTGR7yDTfh0O+PAu+X-P2NwdY4oPMuXrr51awafA@mail.gmail.com/ [2] https://lore.kernel.org/lkml/30fd0e5f2156665c713cf191c5fea9a5548360c0.1609926856.git.viresh.kumar@linaro.org/
Hi Viresh, On Fri, Feb 5, 2021 at 10:55 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > On 05-02-21, 10:41, Geert Uytterhoeven wrote: > > On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > > > Thanks for your patch > > > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > > > will have the usual extension, i.e. .dts, though the blob will have > > > > > > > > Why use .dts and not .dtso for overlays? > > > > Because you originally (until v5) had a single rule for building .dtb > > > > and .dtbo files? > > > > > > I am fine with doing that as well if Rob and David agree to it. Rob > > > did suggest that at one point but we didn't do much about it later on > > > for some reason. > > > > > > FWIW, this will also require a change in the DTC compiler. > > > > Care to explain why? I've been using .dtsi for ages in > > https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays > > I don't see you building them anywhere, they aren't added to the > Makefile ever. What am I missing ? > > actually none of the dtso's were added to any makefile in that branch. E.g. "ARM: dts: Build all overlays if OF_OVERLAY=y"? https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=597ee90971687a45678cca8d16bf624d174a99eb > Anyway, the DTC needs to know how to treat the dtso format and it will > error out currently with unknown format kind of errors. > > Below email [1] have some information on the kind of changes required > here. Also note that we had to do similar changes for dtbo earlier > [2]. > > -- > viresh > > [1] https://lore.kernel.org/lkml/CAK7LNASViCOTGR7yDTfh0O+PAu+X-P2NwdY4oPMuXrr51awafA@mail.gmail.com/ -@ is handled by "kbuild: Enable DT symbols when CONFIG_OF_OVERLAY is used" https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=91e9d998514f3743125a707013a30d5f83054579 > [2] https://lore.kernel.org/lkml/30fd0e5f2156665c713cf191c5fea9a5548360c0.1609926856.git.viresh.kumar@linaro.org/ I never had a need for those changes to dtc. .dtso/.dtbo work fine regardless. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On Fri, Feb 05, 2021 at 11:17:10AM +0100, Geert Uytterhoeven wrote: > Hi Viresh, > > On Fri, Feb 5, 2021 at 10:55 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > On 05-02-21, 10:41, Geert Uytterhoeven wrote: > > > On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > > > > Thanks for your patch > > > > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > > > > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > > > > will have the usual extension, i.e. .dts, though the blob will have > > > > > > > > > > Why use .dts and not .dtso for overlays? > > > > > Because you originally (until v5) had a single rule for building .dtb > > > > > and .dtbo files? > > > > > > > > I am fine with doing that as well if Rob and David agree to it. Rob > > > > did suggest that at one point but we didn't do much about it later on > > > > for some reason. > > > > > > > > FWIW, this will also require a change in the DTC compiler. > > > > > > Care to explain why? I've been using .dtsi for ages in > > > https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays > > > > I don't see you building them anywhere, they aren't added to the > > Makefile ever. What am I missing ? > > > > actually none of the dtso's were added to any makefile in that branch. > > E.g. "ARM: dts: Build all overlays if OF_OVERLAY=y"? > https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=597ee90971687a45678cca8d16bf624d174a99eb > > > Anyway, the DTC needs to know how to treat the dtso format and it will > > error out currently with unknown format kind of errors. > > > > Below email [1] have some information on the kind of changes required > > here. Also note that we had to do similar changes for dtbo earlier > > [2]. > > > > -- > > viresh > > > > [1] https://lore.kernel.org/lkml/CAK7LNASViCOTGR7yDTfh0O+PAu+X-P2NwdY4oPMuXrr51awafA@mail.gmail.com/ > > -@ is handled by "kbuild: Enable DT symbols when CONFIG_OF_OVERLAY is used" > https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=91e9d998514f3743125a707013a30d5f83054579 > > > [2] https://lore.kernel.org/lkml/30fd0e5f2156665c713cf191c5fea9a5548360c0.1609926856.git.viresh.kumar@linaro.org/ > > I never had a need for those changes to dtc. .dtso/.dtbo work fine regardless. I think what Viresh means is dtc won't automatically set the input type to dts if not .dts. We stuck with .dtbo as it's already widely used. I don't know about dtso though. If there's strong consensus to use .dtso, then let's do that. Whatever color for this shed you like. Rob
On Fri, Feb 05, 2021 at 02:55:07PM +0530, Viresh Kumar wrote: > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > Hi Viresh, > > > > Thanks for your patch > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > will have the usual extension, i.e. .dts, though the blob will have > > > > Why use .dts and not .dtso for overlays? > > Because you originally (until v5) had a single rule for building .dtb > > and .dtbo files? > > I am fine with doing that as well if Rob and David agree to it. Rob > did suggest that at one point but we didn't do much about it later on > for some reason. > > FWIW, this will also require a change in the DTC compiler. Not really. It would need a change to automatically recognize that extension, but you can easily work around that by explicitly giving the -I option to specify the input type. -- David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson
Hi Rob, On Fri, Feb 5, 2021 at 10:08 PM Rob Herring <robh@kernel.org> wrote: > On Fri, Feb 05, 2021 at 11:17:10AM +0100, Geert Uytterhoeven wrote: > On Fri, Feb 5, 2021 at 10:55 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > On 05-02-21, 10:41, Geert Uytterhoeven wrote: > > > > On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > > On 05-02-21, 10:02, Geert Uytterhoeven wrote: > > > > > > Thanks for your patch > > > > > > (which I only noticed because it appeared in dt-rh/for-next ;-) > > > > > > > > > > > > On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > > > > > > Add support for building DT overlays (%.dtbo). The overlay's source file > > > > > > > will have the usual extension, i.e. .dts, though the blob will have > > > > > > > > > > > > Why use .dts and not .dtso for overlays? > > > > > > Because you originally (until v5) had a single rule for building .dtb > > > > > > and .dtbo files? > > > > > > > > > > I am fine with doing that as well if Rob and David agree to it. Rob > > > > > did suggest that at one point but we didn't do much about it later on > > > > > for some reason. > > > > > > > > > > FWIW, this will also require a change in the DTC compiler. > I think what Viresh means is dtc won't automatically set the input type > to dts if not .dts. Which is not needed with the kernel build rules. > We stuck with .dtbo as it's already widely used. I don't know about > dtso though. If there's strong consensus to use .dtso, then let's do > that. Whatever color for this shed you like. I did a quick survey of its popularity, and found: 1. .dts => .dtbo https://www.kernel.org/doc/Documentation/devicetree/overlay-notes.txt https://github.com/beagleboard/bb.org-overlays https://www.raspberrypi.org/documentation/configuration/device-tree.md https://www.96boards.org/documentation/consumer/dragonboard/dragonboard410c/guides/dt-overlays.md.html https://learn.adafruit.com/introduction-to-the-beaglebone-black-device-tree/compiling-an-overlay https://developer.toradex.com/knowledge-base/device-tree-overlays 2. .dtso => .dtbo https://github.com/linux4sam/dt-overlay-at91/tree/master/sama5d3_xplained https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays 3. .dtsi => .dtbo https://xilinx-wiki.atlassian.net/wiki/spaces/A/pages/18841645/Solution+Zynq+PL+Programming+With+FPGA+Manager Note that many overlay users haven't migrated to the sugar syntax yet. IMHO the biggest advantage of .dtso over .dts is that you (and the tooling) know which files should be compiled to .dtbo, and which to .dtb. This doesn't matter for obj-$(CONFIG_FOO) += foo.dtbo but it does for options like CONFIG_OF_ALL_DTBS. Having a way to build all overlays will become more important when we will have more generic overlays (capes/hats/wings/...), that can apply to multiple base DTBs. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On 2/5/21 3:08 PM, Rob Herring wrote: > On Fri, Feb 05, 2021 at 11:17:10AM +0100, Geert Uytterhoeven wrote: >> Hi Viresh, >> >> On Fri, Feb 5, 2021 at 10:55 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: >>> On 05-02-21, 10:41, Geert Uytterhoeven wrote: >>>> On Fri, Feb 5, 2021 at 10:25 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: >>>>> On 05-02-21, 10:02, Geert Uytterhoeven wrote: >>>>>> Thanks for your patch >>>>>> (which I only noticed because it appeared in dt-rh/for-next ;-) >>>>>> >>>>>> On Fri, Jan 29, 2021 at 8:31 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: >>>>>>> Add support for building DT overlays (%.dtbo). The overlay's source file >>>>>>> will have the usual extension, i.e. .dts, though the blob will have >>>>>> >>>>>> Why use .dts and not .dtso for overlays? >>>>>> Because you originally (until v5) had a single rule for building .dtb >>>>>> and .dtbo files? >>>>> >>>>> I am fine with doing that as well if Rob and David agree to it. Rob >>>>> did suggest that at one point but we didn't do much about it later on >>>>> for some reason. >>>>> >>>>> FWIW, this will also require a change in the DTC compiler. >>>> >>>> Care to explain why? I've been using .dtsi for ages in >>>> https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/log/?h=topic/renesas-overlays >>> >>> I don't see you building them anywhere, they aren't added to the >>> Makefile ever. What am I missing ? >>> >>> actually none of the dtso's were added to any makefile in that branch. >> >> E.g. "ARM: dts: Build all overlays if OF_OVERLAY=y"? >> https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=597ee90971687a45678cca8d16bf624d174a99eb >> >>> Anyway, the DTC needs to know how to treat the dtso format and it will >>> error out currently with unknown format kind of errors. >>> >>> Below email [1] have some information on the kind of changes required >>> here. Also note that we had to do similar changes for dtbo earlier >>> [2]. >>> >>> -- >>> viresh >>> >>> [1] https://lore.kernel.org/lkml/CAK7LNASViCOTGR7yDTfh0O+PAu+X-P2NwdY4oPMuXrr51awafA@mail.gmail.com/ >> >> -@ is handled by "kbuild: Enable DT symbols when CONFIG_OF_OVERLAY is used" >> https://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-drivers.git/commit/?h=topic/renesas-overlays&id=91e9d998514f3743125a707013a30d5f83054579 >> >>> [2] https://lore.kernel.org/lkml/30fd0e5f2156665c713cf191c5fea9a5548360c0.1609926856.git.viresh.kumar@linaro.org/ >> >> I never had a need for those changes to dtc. .dtso/.dtbo work fine regardless. > > I think what Viresh means is dtc won't automatically set the input type > to dts if not .dts. > > We stuck with .dtbo as it's already widely used. I don't know about > dtso though. If there's strong consensus to use .dtso, then let's do > that. Whatever color for this shed you like. I overlooked this and mistakenly thought that the move to .dtbo also involved changing to .dtso. My bad. My favorite color here is to use .dtso for the source file that will be compiled to create a .dtbo. Linus has already accepted patch 4/6 to 5.12-rc1, so changing to .dtso will require another patch. -Frank > > Rob >
On 24-02-21, 19:32, Frank Rowand wrote: > I overlooked this and mistakenly thought that the move to .dtbo also > involved changing to .dtso. My bad. > > My favorite color here is to use .dtso for the source file that will > be compiled to create a .dtbo. > > Linus has already accepted patch 4/6 to 5.12-rc1, so changing to .dtso > will require another patch. Looks like this is what many people desire, lets do it and make it a standard even if it wasn't followed earlier. What about this ? diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index c430fbb36763..0dbedb61835f 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -337,7 +337,7 @@ $(obj)/%.dtb.S: $(obj)/%.dtb FORCE quiet_cmd_dtc = DTC $@ cmd_dtc = $(HOSTCC) -E $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; \ - $(DTC) -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ + $(DTC) -I dts -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ $(addprefix -i,$(dir $<) $(DTC_INCLUDE)) $(DTC_FLAGS) \ -d $(depfile).dtc.tmp $(dtc-tmp) ; \ cat $(depfile).pre.tmp $(depfile).dtc.tmp > $(depfile) @@ -348,6 +348,9 @@ $(obj)/%.dtb: $(src)/%.dts $(DTC) FORCE $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE $(call if_changed_dep,dtc) +$(obj)/%.dtbo: $(src)/%.dtso $(DTC) FORCE + $(call if_changed_dep,dtc) + overlay-y := $(addprefix $(obj)/, $(overlay-y)) quiet_cmd_fdtoverlay = DTOVL $@ @@ -373,6 +376,9 @@ endef $(obj)/%.dt.yaml: $(src)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE $(call if_changed_rule,dtc,yaml) +$(obj)/%.dt.yaml: $(src)/%.dtso $(DTC) $(DT_TMP_SCHEMA) FORCE + $(call if_changed_rule,dtc,yaml) + dtc-tmp = $(subst $(comma),_,$(dot-target).dts.tmp) # Bzip2 -------------------------8<------------------------- I had to keep the original line as is: $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE to support the unittest stuff as there are no dtso files there. There are few things we can do here: - Don't follow the dtso/dtbo convention for unittest, build files as dtb only and everything will continue to work I suppose as fdtoverlay won't complain. - Keep the above line in Makefile, this doesn't sound right, isn't it ? - Make .dts links for unittest file, maybe from the Makefile itself. - Something else ? -- viresh
Hi Viresh, On Wed, Mar 3, 2021 at 6:21 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > On 24-02-21, 19:32, Frank Rowand wrote: > > I overlooked this and mistakenly thought that the move to .dtbo also > > involved changing to .dtso. My bad. > > > > My favorite color here is to use .dtso for the source file that will > > be compiled to create a .dtbo. > > > > Linus has already accepted patch 4/6 to 5.12-rc1, so changing to .dtso > > will require another patch. > > Looks like this is what many people desire, lets do it and make it a > standard even if it wasn't followed earlier. > > What about this ? Thanks, looks good to me, and works for me, so Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> > --- a/scripts/Makefile.lib > +++ b/scripts/Makefile.lib > @@ -337,7 +337,7 @@ $(obj)/%.dtb.S: $(obj)/%.dtb FORCE > > quiet_cmd_dtc = DTC $@ > cmd_dtc = $(HOSTCC) -E $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; \ > - $(DTC) -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ > + $(DTC) -I dts -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ > $(addprefix -i,$(dir $<) $(DTC_INCLUDE)) $(DTC_FLAGS) \ > -d $(depfile).dtc.tmp $(dtc-tmp) ; \ > cat $(depfile).pre.tmp $(depfile).dtc.tmp > $(depfile) > @@ -348,6 +348,9 @@ $(obj)/%.dtb: $(src)/%.dts $(DTC) FORCE > $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE > $(call if_changed_dep,dtc) > > +$(obj)/%.dtbo: $(src)/%.dtso $(DTC) FORCE > + $(call if_changed_dep,dtc) > + > overlay-y := $(addprefix $(obj)/, $(overlay-y)) > > quiet_cmd_fdtoverlay = DTOVL $@ > @@ -373,6 +376,9 @@ endef > $(obj)/%.dt.yaml: $(src)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE > $(call if_changed_rule,dtc,yaml) > > +$(obj)/%.dt.yaml: $(src)/%.dtso $(DTC) $(DT_TMP_SCHEMA) FORCE I'm wondering if "dt.yaml" should be changed to "dto.yaml" (here and in the existing rule earlier in Makefile.lib), to avoid issues if both foo.dts and foo.dtso exist? Unlikely, but it might happen... > I had to keep the original line as is: > > $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE > > to support the unittest stuff as there are no dtso files there. There > are few things we can do here: > > - Don't follow the dtso/dtbo convention for unittest, build files as > dtb only and everything will continue to work I suppose as > fdtoverlay won't complain. > > - Keep the above line in Makefile, this doesn't sound right, isn't it > ? > > - Make .dts links for unittest file, maybe from the Makefile itself. > > - Something else ? Rename unittest .dts files to .dtso where applicable? Gr{oetje,eeting}s, Geert
On 03-03-21, 11:44, Geert Uytterhoeven wrote: > Hi Viresh, > > On Wed, Mar 3, 2021 at 6:21 AM Viresh Kumar <viresh.kumar@linaro.org> wrote: > > On 24-02-21, 19:32, Frank Rowand wrote: > > > I overlooked this and mistakenly thought that the move to .dtbo also > > > involved changing to .dtso. My bad. > > > > > > My favorite color here is to use .dtso for the source file that will > > > be compiled to create a .dtbo. > > > > > > Linus has already accepted patch 4/6 to 5.12-rc1, so changing to .dtso > > > will require another patch. > > > > Looks like this is what many people desire, lets do it and make it a > > standard even if it wasn't followed earlier. > > > > What about this ? > > Thanks, looks good to me, and works for me, so > Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> > Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Thanks. > > --- a/scripts/Makefile.lib > > +++ b/scripts/Makefile.lib > > @@ -337,7 +337,7 @@ $(obj)/%.dtb.S: $(obj)/%.dtb FORCE > > > > quiet_cmd_dtc = DTC $@ > > cmd_dtc = $(HOSTCC) -E $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; \ > > - $(DTC) -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ > > + $(DTC) -I dts -O $(patsubst .%,%,$(suffix $@)) -o $@ -b 0 \ > > $(addprefix -i,$(dir $<) $(DTC_INCLUDE)) $(DTC_FLAGS) \ > > -d $(depfile).dtc.tmp $(dtc-tmp) ; \ > > cat $(depfile).pre.tmp $(depfile).dtc.tmp > $(depfile) > > @@ -348,6 +348,9 @@ $(obj)/%.dtb: $(src)/%.dts $(DTC) FORCE > > $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE > > $(call if_changed_dep,dtc) > > > > +$(obj)/%.dtbo: $(src)/%.dtso $(DTC) FORCE > > + $(call if_changed_dep,dtc) > > + > > overlay-y := $(addprefix $(obj)/, $(overlay-y)) > > > > quiet_cmd_fdtoverlay = DTOVL $@ > > @@ -373,6 +376,9 @@ endef > > $(obj)/%.dt.yaml: $(src)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE > > $(call if_changed_rule,dtc,yaml) > > > > +$(obj)/%.dt.yaml: $(src)/%.dtso $(DTC) $(DT_TMP_SCHEMA) FORCE > > I'm wondering if "dt.yaml" should be changed to "dto.yaml" (here and in > the existing rule earlier in Makefile.lib), to avoid issues if both foo.dts and > foo.dtso exist? Unlikely, but it might happen... I will let Rob answer that :) > > I had to keep the original line as is: > > > > $(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE > > > > to support the unittest stuff as there are no dtso files there. There > > are few things we can do here: > > > > - Don't follow the dtso/dtbo convention for unittest, build files as > > dtb only and everything will continue to work I suppose as > > fdtoverlay won't complain. > > > > - Keep the above line in Makefile, this doesn't sound right, isn't it > > ? > > > > - Make .dts links for unittest file, maybe from the Makefile itself. > > > > - Something else ? > > Rename unittest .dts files to .dtso where applicable? They are used for some runtime tests, we are reusing them to do this testing as well, so renaming them is out of the question I believe. -- viresh
diff --git a/.gitignore b/.gitignore index d01cda8e1177..bb65fa253e58 100644 --- a/.gitignore +++ b/.gitignore @@ -18,6 +18,7 @@ *.c.[012]*.* *.dt.yaml *.dtb +*.dtbo *.dtb.S *.dwo *.elf diff --git a/Makefile b/Makefile index e0af7a4a5598..d5bc67e523be 100644 --- a/Makefile +++ b/Makefile @@ -1337,6 +1337,9 @@ ifneq ($(dtstree),) %.dtb: include/config/kernel.release scripts_dtc $(Q)$(MAKE) $(build)=$(dtstree) $(dtstree)/$@ +%.dtbo: include/config/kernel.release scripts_dtc + $(Q)$(MAKE) $(build)=$(dtstree) $(dtstree)/$@ + PHONY += dtbs dtbs_install dtbs_check dtbs: include/config/kernel.release scripts_dtc $(Q)$(MAKE) $(build)=$(dtstree) @@ -1816,7 +1819,7 @@ clean: $(clean-dirs) @find $(if $(KBUILD_EXTMOD), $(KBUILD_EXTMOD), .) $(RCS_FIND_IGNORE) \ \( -name '*.[aios]' -o -name '*.ko' -o -name '.*.cmd' \ -o -name '*.ko.*' \ - -o -name '*.dtb' -o -name '*.dtb.S' -o -name '*.dt.yaml' \ + -o -name '*.dtb' -o -name '*.dtbo' -o -name '*.dtb.S' -o -name '*.dt.yaml' \ -o -name '*.dwo' -o -name '*.lst' \ -o -name '*.su' -o -name '*.mod' \ -o -name '.*.d' -o -name '.*.tmp' -o -name '*.mod.c' \ diff --git a/scripts/Makefile.dtbinst b/scripts/Makefile.dtbinst index 50d580d77ae9..ba01f5ba2517 100644 --- a/scripts/Makefile.dtbinst +++ b/scripts/Makefile.dtbinst @@ -29,6 +29,9 @@ quiet_cmd_dtb_install = INSTALL $@ $(dst)/%.dtb: $(obj)/%.dtb $(call cmd,dtb_install) +$(dst)/%.dtbo: $(obj)/%.dtbo + $(call cmd,dtb_install) + PHONY += $(subdirs) $(subdirs): $(Q)$(MAKE) $(dtbinst)=$@ dst=$(patsubst $(obj)/%,$(dst)/%,$@) diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index 213677a5ed33..b00855b247e0 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -86,7 +86,9 @@ extra-$(CONFIG_OF_ALL_DTBS) += $(dtb-) ifneq ($(CHECK_DTBS),) extra-y += $(patsubst %.dtb,%.dt.yaml, $(dtb-y)) +extra-y += $(patsubst %.dtbo,%.dt.yaml, $(dtb-y)) extra-$(CONFIG_OF_ALL_DTBS) += $(patsubst %.dtb,%.dt.yaml, $(dtb-)) +extra-$(CONFIG_OF_ALL_DTBS) += $(patsubst %.dtbo,%.dt.yaml, $(dtb-)) endif # Add subdir path @@ -327,6 +329,9 @@ cmd_dtc = $(HOSTCC) -E $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; $(obj)/%.dtb: $(src)/%.dts $(DTC) FORCE $(call if_changed_dep,dtc) +$(obj)/%.dtbo: $(src)/%.dts $(DTC) FORCE + $(call if_changed_dep,dtc) + DT_CHECKER ?= dt-validate DT_BINDING_DIR := Documentation/devicetree/bindings # DT_TMP_SCHEMA may be overridden from Documentation/devicetree/bindings/Makefile