All of lore.kernel.org
 help / color / mirror / Atom feed
From: Stephen Warren <swarren@wwwdotorg.org>
To: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Cc: "linux-arch@vger.kernel.org" <linux-arch@vger.kernel.org>,
	linux-kernel@vger.kernel.org, Michal Marek <mmarek@suse.cz>,
	Stephen Warren <swarren@nvidia.com>,
	Grant Likely <grant.likely@secretlab.ca>,
	linux-kbuild@vger.kernel.org
Subject: Re: [PATCH] Kbuild: Avoid DTB rebuilds if source files are untouched
Date: Thu, 04 Apr 2013 11:36:14 -0600	[thread overview]
Message-ID: <515DBA0E.6000604@wwwdotorg.org> (raw)
In-Reply-To: <515D0FFB.4030803@synopsys.com>

On 04/03/2013 11:30 PM, Vineet Gupta wrote:
> On 04/03/2013 09:48 PM, Stephen Warren wrote:
>> On 04/03/2013 01:14 AM, Vineet Gupta wrote:
>>> forgot to CC linux-arch
>>>
>>> On 04/03/2013 12:42 PM, Vineet Gupta wrote:
>>>> Currently, for every ARC kernel build I see the following:
>>>>
>>>> --------------->8-----------------
>>>>   DTB    arch/arc/boot/dts/angel4.dtb.S
>>>>   AS      arch/arc/boot/dts/angel4.dtb.o
>>>>   LD      arch/arc/boot/dts/built-in.o
>>>> rm arch/arc/boot/dts/angel4.dtb.S        <-- forces rebuild next iter
>>>>   CHK     kernel/config_data.h
>>>> --------------->8-----------------
>> I assume that's because the file is an intermediate file, and only built
>> due to a chain of build rules, and hence make clean it up itself after
>> the build?
> 
> Indeed - I should have made that explicit in the Changelog.
> 
>>>> +.PRECIOUS: $(obj)/%.dtb.S
>>>> +
>>>>  $(obj)/%.dtb.S: $(obj)/%.dtb
>>>>  	$(call cmd,dt_S_dtb)
>> I'm not sure if .PRECIOUS is correct here. That prevents make from
>> deleting the file if make is CTRL-C'd in the middle of generating it.
>> Couldn't that leave a stale/corrupt file around that'd break the build.
>> Judging by:
>>
>> http://www.gnu.org/software/make/manual/html_node/Special-Targets.html
>>
>> I think .SECONDARY might be a better choice? Does that solve the problem
>> you're seeing?
> 
> Technically .SECONDARY is better - however it doesn't seem to work.

Hmmm. It does for me.

$ make --version
GNU Make 3.81

I hacked the ARM makefiles as follows:

diff --git a/arch/arm/Makefile b/arch/arm/Makefile
index 4737408..70247c6 100644
--- a/arch/arm/Makefile
+++ b/arch/arm/Makefile
@@ -256,6 +256,7 @@ core-y                              +=
arch/arm/kernel/ arch/arm/mm/ arch/arm/common/
 core-y                         += arch/arm/net/
 core-y                         += arch/arm/crypto/
 core-y                         += $(machdirs) $(platdirs)
+core-y                         += arch/arm/boot/dts/

 drivers-$(CONFIG_OPROFILE)      += arch/arm/oprofile/

diff --git a/arch/arm/boot/dts/Makefile b/arch/arm/boot/dts/Makefile
index dedca49..af2202e 100644
--- a/arch/arm/boot/dts/Makefile
+++ b/arch/arm/boot/dts/Makefile
@@ -197,3 +197,5 @@ dtbs: $(addprefix $(obj)/, $(dtb-y))
        $(Q)rm -f $(obj)/../*.dtb

 clean-files := *.dtb
+
+obj-y += tegra20-test.dtb.o

and manually created a tegra20-test.dts. For reasons I didn't bother
investigating, the .dtb.S -> .dtb.o conversion failed with syntax
errors, and because the .dtb.S file was an intermediate file, make rm's
it in this case. That's exactly the issue you're seeing.

I then added the following to Makefile.lib:

diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
index a0ab6d7..fc11a67 100644
--- a/scripts/Makefile.lib
+++ b/scripts/Makefile.lib
@@ -258,6 +258,8 @@ cmd_dt_S_dtb=
        \
        echo '.balign STRUCT_ALIGNMENT';                \
 ) > $@

+.SECONDARY: $(obj)/%.dtb.S
+
 $(obj)/%.dtb.S: $(obj)/%.dtb
        $(call cmd,dt_S_dtb)

and no longer see make rm'ing the .dtb.S file. So, the .SECONDARY is
behaving as expected, and should fix your problem.

One other problem this highlighted: Unless you also mark the generated
.dtb file as PRECIOUS/SECONDARY (or unless that DTB is included in
targets for some reason) then make also rm's that, since it considers it
intermediate:

  DTC     arch/arm/boot/dts/tegra20-test.dtb
  DTB    arch/arm/boot/dts/tegra20-test.dtb.S
  AS      arch/arm/boot/dts/tegra20-test.dtb.o
[error]
rm arch/arm/boot/dts/tegra20-test.dtb

So, your patch would probably need modification to apply the fix to
*.dtb too, since they might also be intermediate?

Perhaps add something like the following to arch/arc/boot/dts/Makefile?

.SECONDARY: $(obj)/$(builtindtb-y).dtb

> Running make with various debug toggles doesn't seem to be helping with why
> .PRECIOUS works but not this.
> That is also likely reason for a bunch of other .PRECIOUS entries in the same file
> but no .SECONDARY.

The existing .PRECIOUS exist to support lexer/yacc files, for which both
the source .l/.y files are checked in, and the generated
%.lex.c_shipped/%.tab.[cl]_shipped are also checked in. In this case,
the "_shipped" files must be marked PRECIOUS, since they're under source
control and hence should never be deleted. I don't believ this same
situation applies to the .dtb.S files you're having problems with, so I
don't think .PRECIOUS is correct for them.

  reply	other threads:[~2013-04-04 17:36 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-04-03  7:12 [PATCH] Kbuild: Avoid DTB rebuilds if source files are untouched Vineet Gupta
2013-04-03  7:12 ` Vineet Gupta
2013-04-03  7:14 ` Vineet Gupta
2013-04-03  7:14   ` Vineet Gupta
2013-04-03 16:18   ` Stephen Warren
2013-04-04  5:30     ` Vineet Gupta
2013-04-04  5:30       ` Vineet Gupta
2013-04-04 17:36       ` Stephen Warren [this message]
2013-04-09 13:37         ` Vineet Gupta
2013-04-09 13:37           ` Vineet Gupta
2013-04-09 14:10         ` Vineet Gupta
2013-04-09 14:10           ` Vineet Gupta
2013-04-12  7:40           ` Vineet Gupta
2013-04-12  7:40             ` Vineet Gupta
2013-04-12 21:52             ` Stephen Warren
2013-04-15 13:59               ` Vineet Gupta
2013-04-15 13:59                 ` Vineet Gupta
2013-04-15 17:12                 ` Stephen Warren
2013-04-16 15:53               ` James Hogan
2013-04-16 15:53                 ` James Hogan
2013-04-16 16:02                 ` James Hogan
2013-04-16 16:02                   ` James Hogan
2013-04-17  4:15                   ` Vineet Gupta
2013-04-17  4:15                     ` Vineet Gupta
2013-04-17  4:13                 ` Vineet Gupta
2013-04-17  4:13                   ` Vineet Gupta
2013-04-17  9:13                   ` James Hogan
2013-04-17  9:13                     ` James Hogan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=515DBA0E.6000604@wwwdotorg.org \
    --to=swarren@wwwdotorg.org \
    --cc=Vineet.Gupta1@synopsys.com \
    --cc=grant.likely@secretlab.ca \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-kbuild@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mmarek@suse.cz \
    --cc=swarren@nvidia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.