git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
To: git@vger.kernel.org
Cc: "Junio C Hamano" <gitster@pobox.com>, "Jeff King" <peff@peff.net>,
	"Andreas Färber" <andreas.faerber@web.de>,
	"Johannes Schindelin" <Johannes.Schindelin@gmx.de>,
	"Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
Subject: [PATCH 4/6] Makefile: make INSTALL_SYMLINKS affect the build directory
Date: Mon, 29 Mar 2021 18:31:42 +0200	[thread overview]
Message-ID: <patch-4.7-1089ca3d184-20210329T162327Z-avarab@gmail.com> (raw)
In-Reply-To: <cover-0.7-00000000000-20210329T162327Z-avarab@gmail.com>

Change the INSTALL_SYMLINKS flag to also affect whether a built-in
like e.g. git-fetch is a symlink or not in git's build directory.

This doesn't matter for anything other than as an aid to developers
who might be confused about their build not matching the installation,
and who'd like to be reminded that e.g. "git-fetch" is a built-in by
"ls" coloring appropriately it as a symlink.

In order to make this work we need to rebuild the relevant programs
when the INSTALL_SYMLINKS flag changes. This also ensures that we'll
install the right thing, we don't want a different "INSTALL_SYMLINKS"
setting during "make all" and "make install" to result in a broken
installation.

We will do the wrong thing here if both the SKIP_DASHED_BUILT_INS and
INSTALL_SYMLINKS are being flipped. But that's not a new bug:

A build with "INSTALL_SYMLINKS=Y SKIP_DASHED_BUILT_INS=" will result
in e.g. "git-fetch" being a symlink. When building again with
"INSTALL_SYMLINKS= SKIP_DASHED_BUILT_INS=Y", only unconditionally
built programs such as "git-upload-pack" will correctly be flipped to
a hardlink, but e.g. "git-fetch" will be left as a symlink.

That's an existing bug (or unexpected behavior) in the
SKIP_DASHED_BUILT_INS flag, not something new being introduced or made
worse here. It's a bit more noticeable now as we might not expect
these now-stale symlinks to be left behind, and "ls" (in some
configurations) will color them prominently.

But we'll still do the right thing on "make install" since we'll
ignore the likes of "git-fetch" there under "SKIP_DASHED_BUILT_INS=Y".
Under "SKIP_DASHED_BUILT_INS=" we'll correctly flip the symlink to a
hardlink or vice-versa if needed before installation.

Still, we should get around to fixing that SKIP_DASHED_BUILT_INS. You
can't reliably set that flag to "Y" for checking whether the tests
rely on the now-skipped dashed built-ins without first running "make
clean" (or knowing you've always been building with
SKIP_DASHED_BUILT_INS=Y).

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
---
 .gitignore |  1 +
 Makefile   | 19 +++++++++++++++++--
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/.gitignore b/.gitignore
index 3dcdb6bb5ab..f90aa21b23b 100644
--- a/.gitignore
+++ b/.gitignore
@@ -5,6 +5,7 @@
 /GIT-BUILD-OPTIONS
 /GIT-CFLAGS
 /GIT-LDFLAGS
+/GIT-LNCPFLAGS
 /GIT-PREFIX
 /GIT-PERL-DEFINES
 /GIT-PERL-HEADER
diff --git a/Makefile b/Makefile
index a4784f28f5b..29d9bade5a8 100644
--- a/Makefile
+++ b/Makefile
@@ -337,6 +337,11 @@ all::
 # NO_INSTALL_HARDLINKS. This will not produce any indirect symlinks, we will
 # always symlink to the final target directly.
 #
+# NO_INSTALL_HARDLINKS which will also use symlinking by indirection
+# within the same directory in some cases, INSTALL_SYMLINKS will
+# always symlink to the final target directly. This option also
+# affects dashed built-ins in the build directory pre-installation.
+#
 # Define NO_CROSS_DIRECTORY_HARDLINKS if you plan to distribute the installed
 # programs as a tar, where bin/ and libexec/ might be on different file systems.
 #
@@ -2197,9 +2202,12 @@ version.sp version.s version.o: EXTRA_CPPFLAGS = \
 		GIT_CEILING_DIRECTORIES="$(CURDIR)/.." \
 		git rev-parse -q --verify HEAD 2>/dev/null)"'
 
+$(BUILT_INS): GIT-LNCPFLAGS
 $(BUILT_INS): git$X
 	$(QUIET_BUILT_IN)$(RM) $@ && \
-	./ln-or-cp.sh $< $@
+	./ln-or-cp.sh \
+		--install-symlinks "$(INSTALL_SYMLINKS)" \
+		$< $@
 
 config-list.h: generate-configlist.sh
 
@@ -2548,9 +2556,12 @@ git-http-push$X: http.o http-push.o GIT-LDFLAGS $(GITLIBS)
 		$(CURL_LIBCURL) $(EXPAT_LIBEXPAT) $(LIBS) && \
 	mv $@+ $@
 
+$(REMOTE_CURL_ALIASES): GIT-LNCPFLAGS
 $(REMOTE_CURL_ALIASES): $(REMOTE_CURL_PRIMARY)
 	$(QUIET_LNCP)$(RM) $@ && \
-	./ln-or-cp.sh $< $@
+	./ln-or-cp.sh \
+		--install-symlinks "$(INSTALL_SYMLINKS)" \
+		$< $@
 
 $(REMOTE_CURL_PRIMARY): remote-curl.o http.o http-walker.o GIT-LDFLAGS $(GITLIBS)
 	$(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@+ $(ALL_LDFLAGS) $(filter %.o,$^) \
@@ -2742,6 +2753,10 @@ GIT-LDFLAGS: FORCE
 		echo "$$FLAGS" >GIT-LDFLAGS; \
             fi
 
+GIT-LNCPFLAGS: FORCE
+	@echo INSTALL_SYMLINKS=\''$(subst ','\'',$(subst ','\'',$(INSTALL_SYMLINKS)))'\' >$@+
+	@if cmp $@+ $@ >/dev/null 2>&1; then $(RM) $@+; else mv $@+ $@; fi
+
 # We need to apply sq twice, once to protect from the shell
 # that runs GIT-BUILD-OPTIONS, and then again to protect it
 # and the first level quoting from the shell that runs "echo".
-- 
2.31.1.461.gd47399f6574


  parent reply	other threads:[~2021-03-29 16:33 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-07 13:20 [PATCH] Makefile: generate 'git' as 'cc [...] -o git+ && mv git+ git' Ævar Arnfjörð Bjarmason
2021-03-07 20:41 ` Junio C Hamano
2021-03-08 12:38   ` Ævar Arnfjörð Bjarmason
2021-03-08 17:21     ` Junio C Hamano
2021-03-08 18:26     ` Jeff King
2021-03-29 16:20 ` [PATCH v2 0/5] Makefile: don't die on AIX with open ./git Ævar Arnfjörð Bjarmason
2021-03-29 16:20   ` [PATCH v2 1/5] Makefile: rename objects in-place, don't clobber Ævar Arnfjörð Bjarmason
2021-03-29 18:21     ` Junio C Hamano
2021-03-29 18:26       ` Junio C Hamano
2021-03-29 23:24       ` Ævar Arnfjörð Bjarmason
2021-03-30  0:21         ` Junio C Hamano
2021-03-31 14:17           ` Ævar Arnfjörð Bjarmason
2021-03-31 18:50             ` Junio C Hamano
2021-03-29 16:20   ` [PATCH v2 2/5] Makefile: rename scripts " Ævar Arnfjörð Bjarmason
2021-03-29 18:40     ` Junio C Hamano
2021-03-29 23:28       ` Ævar Arnfjörð Bjarmason
2021-03-29 16:20   ` [PATCH v2 3/5] Makefile: don't needlessly "rm $@ $@+" before "mv $@+ $@" Ævar Arnfjörð Bjarmason
2021-03-29 18:46     ` Junio C Hamano
2021-03-29 16:20   ` [PATCH v2 4/5] Makefile: add the ".DELETE_ON_ERROR" flag Ævar Arnfjörð Bjarmason
2021-03-29 18:51     ` Junio C Hamano
2021-03-29 23:31       ` Ævar Arnfjörð Bjarmason
2021-03-29 23:58         ` Junio C Hamano
2021-03-30 15:11         ` Jeff King
2021-03-30 18:36           ` Junio C Hamano
2021-03-31  6:58             ` Jeff King
2021-03-31 18:36               ` Junio C Hamano
2021-03-31 22:29                 ` Johannes Schindelin
2021-03-29 16:20   ` [PATCH v2 5/5] Makefile: don't "rm configure" before generating it Ævar Arnfjörð Bjarmason
2021-03-29 16:31   ` [PATCH 0/6] Makefile: make non-symlink & non-hardlink install sane Ævar Arnfjörð Bjarmason
2021-03-29 16:31     ` [PATCH 1/6] Makefile: symlink the same way under "symlinks" and "no hardlinks" Ævar Arnfjörð Bjarmason
2021-03-29 22:17       ` Junio C Hamano
2021-03-29 16:31     ` [PATCH 2/6] Makefile: begin refactoring out "ln || ln -s || cp" pattern Ævar Arnfjörð Bjarmason
2021-03-29 22:20       ` Junio C Hamano
2021-03-29 16:31     ` [PATCH 3/6] Makefile: refactor " Ævar Arnfjörð Bjarmason
2021-03-29 22:24       ` Junio C Hamano
2021-03-30 15:20         ` Jeff King
2021-03-30 18:36           ` Junio C Hamano
2021-03-29 16:31     ` Ævar Arnfjörð Bjarmason [this message]
2021-03-29 22:31       ` [PATCH 4/6] Makefile: make INSTALL_SYMLINKS affect the build directory Junio C Hamano
2021-03-31 14:04         ` Ævar Arnfjörð Bjarmason
2021-03-29 16:31     ` [PATCH 5/6] Makefile: use "ln -f" instead of "rm && ln" Ævar Arnfjörð Bjarmason
2021-03-29 22:46       ` Junio C Hamano
2021-03-29 16:31     ` [PATCH 6/6] Makefile: add a INSTALL_FALLBACK_LN_CP mode Ævar Arnfjörð Bjarmason
2021-03-29 22:53       ` Junio C Hamano
2021-03-31 14:03         ` Ævar Arnfjörð Bjarmason
2021-03-31 18:45           ` Junio C Hamano
2021-03-31 19:01             ` Ævar Arnfjörð Bjarmason
2021-03-29 23:08     ` [PATCH 0/6] Makefile: make non-symlink & non-hardlink install sane Junio C Hamano

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=patch-4.7-1089ca3d184-20210329T162327Z-avarab@gmail.com \
    --to=avarab@gmail.com \
    --cc=Johannes.Schindelin@gmx.de \
    --cc=andreas.faerber@web.de \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=peff@peff.net \
    /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;
as well as URLs for NNTP newsgroup(s).