On Mon, 12 Sep 2022 10:01:52 +0200 Sascha Hauer <s.hauer@xxxxxxxxxxxxxx> wrote: Hi Sascha! > On Wed, Sep 07, 2022 at 11:21:18AM +0300, Antony Pavlov wrote: > > Hi Everyone! > > > > During MIPS ath79_defconfig build I have a 'missing whitespace after the macro name' warning: > > > > AS [P] arch/mips/dts/ar9331_tl_mr3020.dtb.pbl.o > > DTC arch/mips/dts/ar9344-tl-wdr4300-v1.7.dtb > > <stdin>:1:9: warning: missing whitespace after the macro name > > XZKERN arch/mips/dts/ar9344-tl-wdr4300-v1.7.dtb.z > > > > E.g. see https://gitlab.com/frantony/barebox/-/jobs/2969826747#L47 > > > > The reason is the scripts/Makefile.lib generates a C macro with the point symbol in the macro name because the arch/mips/dts/ar9344-tl-wdr4300-v1.7.dts file name contains the point symbol before '.dts', as a result we have: > > > > #define ar9344_tl_wdr4300_v1.7_dts 1 > > > > e.g. > > > > barebox$ grep -RHn -o "define ar9344_tl_wdr4300_v1.* 1" . 2>/dev/null > > ./arch/mips/dts/.ar9344-tl-wdr4300-v1.7.dtb.cmd:1:define ar9344_tl_wdr4300_v1.7_dts 1 > > > > cmd_dtc in scripts/Makefile.lib substitutes the '-' symbols with the '_' symbols but do nothing with other unwanted C preprocessor macro name symbols. > > > > It looks like the linux kernel has no problems with extra point symbols in dts file names, there are several files with extra dot in dts: > > > > barebox$ find dts/ -iname '*.*.dts' | wc -l > > 33 > > > > So we have to fix Makefile.lib. > > > > This simple patch fixes the warning problem: > > > > diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib > > index 16308497b84..2f79656c1e9 100644 > > --- a/scripts/Makefile.lib > > +++ b/scripts/Makefile.lib > > @@ -363,7 +363,7 @@ $(obj)/%.dtb.z: $(obj)/%.dtb FORCE > > dts-frags = $(subst $(quote),,$(CONFIG_EXTERNAL_DTS_FRAGMENTS)) > > quiet_cmd_dtc = DTC $@ > > # For compatibility between make 4.2 and 4.3 > > -cmd_dtc = /usr/bin/env echo -e '$(pound)define $(subst -,_,$(*F))_dts 1\n'$(foreach f,$< $(dts-frags),'$(pound)include "$(f)"\n') | \ > > +cmd_dtc = /usr/bin/env echo -e '$(pound)define $(subst -,_,$(subst .,_,$(*F)))_dts 1\n'$(foreach f,$< $(dts-frags),'$(pound)include "$(f)"\n') | \ > > $(CPP) $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) - ; \ > > $(objtree)/scripts/dtc/dtc -O dtb -o $@ -b 0 \ > > -i $(srctree)/arch/$(SRCARCH)/dts $(DTC_FLAGS) \ > > > > I suppose that this simple patch may lead to some undesirable side effects. > > One side effect is that this gets even less readable. Yes, you are right :) > Another one would be that two dts filenames which only differ in the > usage of '.' and '_' would result in the same define, but I think that > case is negligible as this define is unused in barebox itself. It could ^^^^^^^^^^^^^^^^^^^^^^^^ > be used by external dts fragments passed in via CONFIG_EXTERNAL_DTS_FRAGMENTS. > > Other than that, what side effects are you afraid of? I have an idea that someone relies on these *_dts macros. Now I see (as you have noted above) that there is no macro users in mainline source tree: $ git grep -w .*_dts common/Kconfig: #ifdef foo_board_dts scripts/Makefile.lib:cmd_dtc = /usr/bin/env echo -e '$(pound)define $(subst -,_,$(*F))_dts 1\n'$(foreach f,$< $(dts-frags),'$(pound)include "$(f)"\n') | \ -- Best regards, Antony Pavlov
