From: NeilBrown <neilb@suse.com>
To: Christoph Hellwig <hch@infradead.org>
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>,
Michal Marek <michal.lkml@markovi.net>,
linux-kernel@vger.kernel.org, linux-kbuild@vger.kernel.org
Subject: Re: [RFC PATCH 0/5] kbuild: build modules from code in multiple directories.
Date: Tue, 19 Jun 2018 14:05:23 +1000 [thread overview]
Message-ID: <87po0ne7zw.fsf@notabene.neil.brown.name> (raw)
In-Reply-To: <20180618082017.GA4752@infradead.org>
[-- Attachment #1: Type: text/plain, Size: 6271 bytes --]
On Mon, Jun 18 2018, Christoph Hellwig wrote:
> On Mon, Jun 18, 2018 at 02:55:20PM +1000, NeilBrown wrote:
>> This set of patches makes it possible to build a module from
>> code in multiple directories without needing to list files from one
>> directory in the Makefile of another directory.
>>
>> The code was developed for lustre (which is now out-of-tree :-( ) but
>> can be useful elsewhere, such as for xfs and btrfs and others.
>>
>> In fs/xfs/Makefile the section:
>>
>> xfs-y += $(addprefix libxfs/, \
>> xfs_ag.o \
>> xfs_alloc.o \
>> .....
>>
>> could become
>>
>> xfs-y += libxfs/
>>
>> and then in fs/xfs/libxfs/Makefile we would have
>>
>> modobj-$(CONFIG_XFS_FS) += xfs_ag.o \
>> xfs_alloc.o \
>> .....
>>
>> A similar process could move filenames for scrub/* from the
>> fs/xfs/Makefile to fs/xfs/scrub/Makefile
>
> How about you actually convert it as an example?
Sure ... found a bug while testing it. Thanks :-)
From ac7953b4ba6d9a2a69dab84dd772aafca38d8377 Mon Sep 17 00:00:00 2001
From: NeilBrown <neilb@suse.com>
Date: Tue, 19 Jun 2018 13:59:16 +1000
Subject: [PATCH] kbuild/xfs: example modobj-m conversion
This is a demonstration patch to show how
xfs can be changed to make use of the proposed modobj-m=
functionality, should the xfs developers want that.
Signed-off-by: NeilBrown <neilb@suse.com>
---
fs/xfs/Makefile | 78 ++------------------------------------------------
fs/xfs/libxfs/Makefile | 43 ++++++++++++++++++++++++++++
fs/xfs/scrub/Makefile | 29 +++++++++++++++++++
3 files changed, 74 insertions(+), 76 deletions(-)
create mode 100644 fs/xfs/libxfs/Makefile
create mode 100644 fs/xfs/scrub/Makefile
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 2f3f75a7f180..0ba854045fe9 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -15,47 +15,7 @@ obj-$(CONFIG_XFS_FS) += xfs.o
xfs-y += xfs_trace.o
# build the libxfs code first
-xfs-y += $(addprefix libxfs/, \
- xfs_ag.o \
- xfs_alloc.o \
- xfs_alloc_btree.o \
- xfs_attr.o \
- xfs_attr_leaf.o \
- xfs_attr_remote.o \
- xfs_bit.o \
- xfs_bmap.o \
- xfs_bmap_btree.o \
- xfs_btree.o \
- xfs_da_btree.o \
- xfs_da_format.o \
- xfs_defer.o \
- xfs_dir2.o \
- xfs_dir2_block.o \
- xfs_dir2_data.o \
- xfs_dir2_leaf.o \
- xfs_dir2_node.o \
- xfs_dir2_sf.o \
- xfs_dquot_buf.o \
- xfs_ialloc.o \
- xfs_ialloc_btree.o \
- xfs_iext_tree.o \
- xfs_inode_fork.o \
- xfs_inode_buf.o \
- xfs_log_rlimit.o \
- xfs_ag_resv.o \
- xfs_rmap.o \
- xfs_rmap_btree.o \
- xfs_refcount.o \
- xfs_refcount_btree.o \
- xfs_sb.o \
- xfs_symlink_remote.o \
- xfs_trans_resv.o \
- xfs_types.o \
- )
-# xfs_rtbitmap is shared with libxfs
-xfs-$(CONFIG_XFS_RT) += $(addprefix libxfs/, \
- xfs_rtbitmap.o \
- )
+xfs-y += libxfs/
# highlevel code
xfs-y += xfs_aops.o \
@@ -127,38 +87,4 @@ xfs-$(CONFIG_SYSCTL) += xfs_sysctl.o
xfs-$(CONFIG_COMPAT) += xfs_ioctl32.o
xfs-$(CONFIG_EXPORTFS_BLOCK_OPS) += xfs_pnfs.o
-# online scrub/repair
-ifeq ($(CONFIG_XFS_ONLINE_SCRUB),y)
-
-# Tracepoints like to blow up, so build that before everything else
-
-xfs-y += $(addprefix scrub/, \
- trace.o \
- agheader.o \
- alloc.o \
- attr.o \
- bmap.o \
- btree.o \
- common.o \
- dabtree.o \
- dir.o \
- ialloc.o \
- inode.o \
- parent.o \
- refcount.o \
- rmap.o \
- scrub.o \
- symlink.o \
- )
-
-xfs-$(CONFIG_XFS_RT) += scrub/rtbitmap.o
-xfs-$(CONFIG_XFS_QUOTA) += scrub/quota.o
-
-# online repair
-ifeq ($(CONFIG_XFS_ONLINE_REPAIR),y)
-xfs-y += $(addprefix scrub/, \
- agheader_repair.o \
- repair.o \
- )
-endif
-endif
+xfs-$(CONFIG_XFS_ONLINE_SCRUB) += scrub/
diff --git a/fs/xfs/libxfs/Makefile b/fs/xfs/libxfs/Makefile
new file mode 100644
index 000000000000..9b90c289bdda
--- /dev/null
+++ b/fs/xfs/libxfs/Makefile
@@ -0,0 +1,43 @@
+# SPDX-License-Identifier: GPL-2.0
+
+ccflags-y += -I$(src)/..
+
+# xfs_rtbitmap is shared with libxfs
+xobj-$(CONFIG_XFS_RT) += xfs_rtbitmap.o
+
+modobj-$(CONFIG_XFS_FS) += xfs_ag.o \
+ xfs_alloc.o \
+ xfs_alloc_btree.o \
+ xfs_attr.o \
+ xfs_attr_leaf.o \
+ xfs_attr_remote.o \
+ xfs_bit.o \
+ xfs_bmap.o \
+ xfs_bmap_btree.o \
+ xfs_btree.o \
+ xfs_da_btree.o \
+ xfs_da_format.o \
+ xfs_defer.o \
+ xfs_dir2.o \
+ xfs_dir2_block.o \
+ xfs_dir2_data.o \
+ xfs_dir2_leaf.o \
+ xfs_dir2_node.o \
+ xfs_dir2_sf.o \
+ xfs_dquot_buf.o \
+ xfs_ialloc.o \
+ xfs_ialloc_btree.o \
+ xfs_iext_tree.o \
+ xfs_inode_fork.o \
+ xfs_inode_buf.o \
+ xfs_log_rlimit.o \
+ xfs_ag_resv.o \
+ xfs_rmap.o \
+ xfs_rmap_btree.o \
+ xfs_refcount.o \
+ xfs_refcount_btree.o \
+ xfs_sb.o \
+ xfs_symlink_remote.o \
+ xfs_trans_resv.o \
+ xfs_types.o \
+ $(xobj-y)
diff --git a/fs/xfs/scrub/Makefile b/fs/xfs/scrub/Makefile
new file mode 100644
index 000000000000..f91818a0e6ee
--- /dev/null
+++ b/fs/xfs/scrub/Makefile
@@ -0,0 +1,29 @@
+# SPDX-License-Identifier: GPL-2.0
+
+# Tracepoints like to blow up, so build that before everything else
+
+ccflags-y += -I$(src)/..
+ccflags-y += -I$(src)/../libxfs
+
+xobj-$(CONFIG_XFS_RT) += rtbitmap.o
+xobj-$(CONFIG_XFS_QUOTA) += quota.o
+xobj-$(CONFIG_XFS_ONLINE_REPAIR)+= agheader_repair.o \
+ repair.o
+
+modobj-$(CONFIG_XFS_FS) += trace.o \
+ agheader.o \
+ alloc.o \
+ attr.o \
+ bmap.o \
+ btree.o \
+ common.o \
+ dabtree.o \
+ dir.o \
+ ialloc.o \
+ inode.o \
+ parent.o \
+ refcount.o \
+ rmap.o \
+ scrub.o \
+ symlink.o \
+ $(xobj-y)
--
2.14.0.rc0.dirty
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]
next prev parent reply other threads:[~2018-06-19 4:05 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-18 4:55 NeilBrown
2018-06-18 4:55 ` [PATCH 4/5] kbuild: disable KBUILD_MODNAME when building for mod.a NeilBrown
2018-06-27 5:52 ` Masahiro Yamada
2018-07-03 22:14 ` NeilBrown
2018-07-04 12:08 ` Masahiro Yamada
2018-07-04 21:54 ` NeilBrown
2018-07-05 9:06 ` Masahiro Yamada
2018-07-05 23:03 ` NeilBrown
2018-06-18 4:55 ` [PATCH 1/5] kbuild: detect directories in components of a module NeilBrown
2018-06-18 4:55 ` [PATCH 2/5] kbuild: treat a directory listed in a composite object as foo/mod.a NeilBrown
2018-06-18 9:14 ` kbuild test robot
2018-06-18 22:30 ` NeilBrown
2018-06-18 4:55 ` [PATCH 5/5] kbuild: Add documentation for modobj-m NeilBrown
2018-06-18 4:55 ` [PATCH 3/5] kbuild: support building of per-directory mod.a NeilBrown
2018-06-19 3:57 ` [PATCH 3/5 - v2] " NeilBrown
2018-06-18 8:20 ` [RFC PATCH 0/5] kbuild: build modules from code in multiple directories Christoph Hellwig
2018-06-19 4:05 ` NeilBrown [this message]
2018-06-19 4:47 ` Christoph Hellwig
2018-06-19 5:03 ` Darrick J. Wong
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=87po0ne7zw.fsf@notabene.neil.brown.name \
--to=neilb@suse.com \
--cc=hch@infradead.org \
--cc=linux-kbuild@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michal.lkml@markovi.net \
--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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®