All of lore.kernel.org
 help / color / mirror / Atom feed
From: Matthias Maennich <maennich@google.com>
To: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: Jessica Yu <jeyu@kernel.org>,
	Linux Kbuild mailing list <linux-kbuild@vger.kernel.org>,
	Matthias Kaehlcke <mka@chromium.org>,
	Jonathan Corbet <corbet@lwn.net>,
	Michal Marek <michal.lkml@markovi.net>,
	"open list:DOCUMENTATION" <linux-doc@vger.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 2/4] modpost: dump missing namespaces into a single modules.nsdeps file
Date: Wed, 6 Nov 2019 15:24:54 +0000	[thread overview]
Message-ID: <20191106152454.GA1243@google.com> (raw)
In-Reply-To: <CAK7LNASCmSUuqLyJhZW+3yrGk1KTPxA-_0x86N=Y7A5psCVUSg@mail.gmail.com>

On Thu, Oct 31, 2019 at 08:53:25PM +0900, Masahiro Yamada wrote:
>On Thu, Oct 31, 2019 at 8:20 PM Jessica Yu <jeyu@kernel.org> wrote:
>>
>> +++ Masahiro Yamada [29/10/19 21:38 +0900]:
>> >The modpost, with the -d option given, generates per-module .ns_deps
>> >files.
>> >
>> >Kbuild generates per-module .mod files to carry module information.
>> >This is convenient because Make handles multiple jobs when the -j
>> >option is given.
>> >
>> >On the other hand, the modpost always runs as a single thread.
>> >I do not see a strong reason to produce separate .ns_deps files.
>> >
>> >This commit changes the modpost to generate just one file,
>> >modules.nsdeps, each line of which has the following format:
>> >
>> >  <module_name>: <list of missing namespaces>
>> >
>> >Please note it contains *missing* namespaces instead of required ones.
>> >So, modules.nsdeps is empty if the namespace dependency is all good.
>> >
>> >This will work more efficiently because spatch will no longer process
>> >already imported namespaces. I removed the '(if needed)' from the
>> >nsdeps log since spatch is invoked only when needed.
>> >
>> >This also solved the stale .ns_deps files problem reported by
>> >Jessica Yu:
>> >
>> >  https://lkml.org/lkml/2019/10/28/467
>> >
>> >While I was here, I improved the modpost code a little more;
>> >I freed ns_deps_bus.p because buf_write() allocates memory.
>> >
>> >Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
>> >---
>> >
>> > .gitignore               |  2 +-
>> > Documentation/dontdiff   |  1 +
>> > Makefile                 |  4 ++--
>> > scripts/Makefile.modpost |  2 +-
>> > scripts/mod/modpost.c    | 44 +++++++++++++++++-----------------------
>> > scripts/mod/modpost.h    |  4 ++--
>> > scripts/nsdeps           | 21 +++++++++----------
>> > 7 files changed, 36 insertions(+), 42 deletions(-)
>> >
>> >diff --git a/.gitignore b/.gitignore
>> >index 70580bdd352c..72ef86a5570d 100644
>> >--- a/.gitignore
>> >+++ b/.gitignore
>> >@@ -32,7 +32,6 @@
>> > *.lzo
>> > *.mod
>> > *.mod.c
>> >-*.ns_deps
>> > *.o
>> > *.o.*
>> > *.patch
>> >@@ -61,6 +60,7 @@ modules.order
>> > /System.map
>> > /Module.markers
>> > /modules.builtin.modinfo
>> >+/modules.nsdeps
>> >
>> > #
>> > # RPM spec file (make rpm-pkg)
>> >diff --git a/Documentation/dontdiff b/Documentation/dontdiff
>> >index 9f4392876099..72fc2e9e2b63 100644
>> >--- a/Documentation/dontdiff
>> >+++ b/Documentation/dontdiff
>> >@@ -179,6 +179,7 @@ mkutf8data
>> > modpost
>> > modules.builtin
>> > modules.builtin.modinfo
>> >+modules.nsdeps
>> > modules.order
>> > modversions.h*
>> > nconf
>> >diff --git a/Makefile b/Makefile
>> >index 0ef897fd9cfd..1e3f307bd49b 100644
>> >--- a/Makefile
>> >+++ b/Makefile
>> >@@ -1356,7 +1356,7 @@ endif # CONFIG_MODULES
>> >
>> > # Directories & files removed with 'make clean'
>> > CLEAN_DIRS  += include/ksym
>> >-CLEAN_FILES += modules.builtin.modinfo
>> >+CLEAN_FILES += modules.builtin.modinfo modules.nsdeps
>>
>> Hmm, I tried to run `make -C path/to/kernel/src M=$(PWD) clean` for a test
>> external module, but it didn't remove modules.nsdeps for me. I declared a
>> MODULE namespace for testing purposes.
>>
>> $ make
>> make -C /dev/shm/linux M=/tmp/ppyu/test-module
>> make[1]: Entering directory '/dev/shm/linux'
>>   AR      /tmp/ppyu/test-module/built-in.a
>>   CC [M]  /tmp/ppyu/test-module/test1.o
>>   CC [M]  /tmp/ppyu/test-module/test2.o
>>   LD [M]  /tmp/ppyu/test-module/test.o
>>   Building modules, stage 2.
>>   MODPOST 1 modules
>> WARNING: module test uses symbol try_module_get from namespace MODULE, but does not import it.
>>   CC [M]  /tmp/ppyu/test-module/test.mod.o
>>   LD [M]  /tmp/ppyu/test-module/test.ko
>> make[1]: Leaving directory '/dev/shm/linux'
>>
>> Then I make nsdeps:
>>
>> make -C /dev/shm/linux M=/tmp/ppyu/test-module nsdeps
>> make[1]: Entering directory '/dev/shm/linux'
>>   Building modules, stage 2.
>>   MODPOST 1 modules
>> WARNING: module test uses symbol try_module_get from namespace MODULE, but does not import it.
>> Adding namespace MODULE to module /tmp/ppyu/test-module/test.ko.
>> --- /tmp/ppyu/test-module/test1.c
>> +++ /tmp/cocci-output-3696-c1f8b3-test1.c
>> @@ -38,3 +38,4 @@ static void __exit hello_cleanup(void)
>>  module_init(hello_init);
>>  module_exit(hello_cleanup);
>>  MODULE_LICENSE("GPL");
>> +MODULE_IMPORT_NS(MODULE);
>> make[1]: Leaving directory '/dev/shm/linux'
>> $ cat modules.nsdeps
>> /tmp/ppyu/test-module/test.ko: MODULE
>>
>> Looks good so far, then I try make clean:
>>
>> $ make clean
>> make -C /dev/shm/linux M=/tmp/ppyu/test-module clean
>> make[1]: Entering directory '/dev/shm/linux'
>>   CLEAN   /tmp/ppyu/test-module/Module.symvers
>> make[1]: Leaving directory '/dev/shm/linux'
>> $ ls
>> Makefile  modules.nsdeps  test1.c  test2.c
>>
>> But modules.nsdeps is still there.
>>
>
>Good catch!
>
>I forgot to take care of it for external module builds.
>
>The following should work. I will fold it in 3/4.
>
>
>
>
>diff --git a/Makefile b/Makefile
>index 79be70bf2899..6035702803eb 100644
>--- a/Makefile
>+++ b/Makefile
>@@ -1619,7 +1619,7 @@ _emodinst_post: _emodinst_
>        $(call cmd,depmod)
>
> clean-dirs := $(KBUILD_EXTMOD)
>-clean: rm-files := $(KBUILD_EXTMOD)/Module.symvers
>+clean: rm-files := $(KBUILD_EXTMOD)/Module.symvers $(KBUILD_EXTMOD)/modules.nsdeps

Reviewed-by: Matthias Maennich <maennich@google.com>
Tested-by: Matthias Maennich <maennich@google.com>

Thanks for this improvement!

Cheers,
Matthias

>
> PHONY += /
> /:
>
>
>
>-- 
>Best Regards
>Masahiro Yamada

  reply	other threads:[~2019-11-06 15:25 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-29 12:38 [PATCH 0/4] More nsdeps improvements Masahiro Yamada
2019-10-29 12:38 ` [PATCH 1/4] modpost: do not invoke extra modpost for nsdeps Masahiro Yamada
2019-11-05 16:37   ` Matthias Maennich
2019-10-29 12:38 ` [PATCH 2/4] modpost: dump missing namespaces into a single modules.nsdeps file Masahiro Yamada
2019-10-30 20:11   ` Jessica Yu
2019-10-31 11:20   ` Jessica Yu
2019-10-31 11:53     ` Masahiro Yamada
2019-11-06 15:24       ` Matthias Maennich [this message]
2019-10-29 12:38 ` [PATCH 3/4] scripts/nsdeps: support nsdeps for external module builds Masahiro Yamada
2019-10-30 17:02   ` Jessica Yu
2019-11-06 16:12   ` Matthias Maennich
2019-10-29 12:38 ` [PATCH 4/4] mospost: remove unneeded local variable in contains_namespace() Masahiro Yamada
2019-11-05 13:47   ` Matthias Maennich
2019-11-06 15:39   ` Masahiro Yamada
2019-10-30 20:14 ` [PATCH 0/4] More nsdeps improvements Jessica Yu

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=20191106152454.GA1243@google.com \
    --to=maennich@google.com \
    --cc=corbet@lwn.net \
    --cc=jeyu@kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kbuild@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=michal.lkml@markovi.net \
    --cc=mka@chromium.org \
    --cc=yamada.masahiro@socionext.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.