linux-kernel - Re: [PATCH V5 1/2] kbuild: centralize .dts->.dtb rule

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <50A5350A.3070109@wwwdotorg.org>
Date:	Thu, 15 Nov 2012 11:31:38 -0700
From:	Stephen Warren <swarren@...dotorg.org>
To:	Grant Likely <grant.likely@...retlab.ca>
CC:	linux-arch@...r.kernel.org, linux-mips@...ux-mips.org,
	Stephen Warren <swarren@...dia.com>,
	devicetree-discuss@...ts.ozlabs.org, linux-kernel@...r.kernel.org,
	Sam Ravnborg <sam@...nborg.org>
Subject: Re: [PATCH V5 1/2] kbuild: centralize .dts->.dtb rule

On 11/15/2012 11:20 AM, Grant Likely wrote:
> On Thu, Nov 15, 2012 at 6:09 PM, Stephen Warren <swarren@...dotorg.org> wrote:
>> On 11/15/2012 04:51 AM, Grant Likely wrote:
>>> Grant Likely wrote:
>>>> Or how about: I could pick up the patch with only the MIPS hunk and
>>>> every other user can be fixed up independently to use the new rule.
>>>
>>> Here's a trial patch to fix up ARM. Does this look correct? This patch
>>> depends on the generic dtb build rule already being applied.
>>
>> I think the patch looks OK technically, except for one minor comment below.
>>
>> One issue with this patch is that it moves *.dts from arch/arm/boot to
>> arch/arm/boot/dts, which means everyone has to adjust their scripts/...
>> that package/install/... the kernel. I guess it's an easy change for
>> people to make, but could easily catch people unawares if they do
>> incremental builds so that arch/arm/boot/*.dtb still exists but is stale.
> 
> True. We could temporarily remove or rename if the same file exists in
> the directory below to help people catch that problem. I really would
> like to clean up that build rule to be consistent though.
> 
> The other option is to move all the .dts files into the boot
> directory, but I don't think that is a good idea at all.

Maybe we can just add "rm *.dtb" to the following rules in
boot/Makefile, before calling the child make?

%.dtb: scripts
-	$(Q)$(MAKE) $(build)=$(boot) MACHINE=$(MACHINE) $(boot)/$@
+	$(Q)$(MAKE) $(build)=$(boot)/dts MACHINE=$(MACHINE) $(boot)/dts/$@

 dtbs: scripts
-	$(Q)$(MAKE) $(build)=$(boot) MACHINE=$(MACHINE) $(boot)/$@
+	$(Q)$(MAKE) $(build)=$(boot)/dts MACHINE=$(MACHINE) dtbs

>>> +targets += dtbs
>>
>> Doesn't that make the "dtbs" target always run by default? Perhaps
>> that's reasonable though, and doesn't actually affect anything since the
>> make command for this directory always specifies an explicit target?
>>
>> Or, was that meant to be the following that got removed from ../Makefile?
>>
>> targets += $(dtb-y)
> 
> Yes it is supposed to be the same thing. Doesn't it effectively do the
> same since dtbs depends on $(dtb-y)?

Ah, I think so yes.

I guess anyway that $(targets) is presumably ignored if an explicit
build target is requested from make.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/