From: sashiko-bot@kernel.org
To: "Tomas Glozar" <tglozar@redhat.com>
Cc: linux-perf-users@vger.kernel.org
Subject: Re: [PATCH 1/3] rtla: Add libsubcmd dependency
Date: Sat, 25 Apr 2026 01:37:39 +0000 [thread overview]
Message-ID: <20260425013739.BF4C4C19425@smtp.kernel.org> (raw)
In-Reply-To: <20260320150651.51057-2-tglozar@redhat.com>
Sashiko AI review found 2 potential issue(s):
- [Low] The variable `LIBSUBCMD_INCLUDES` is incorrectly used as both a compiler flag and a Make target, causing unnecessary sub-make executions.
- [Low] The manual compilation rules for `string.o` and `str_error_r.o` bypass the standard `fixdep` dependency tracking mechanism.
--
commit 61af24c97cebeb1e051afb96f7fccb2a5c359a72
Author: Tomas Glozar <tglozar@redhat.com>
rtla: Add libsubcmd dependency
This commit prepares RTLA to migrate to libsubcmd by building and linking
the library alongside the RTLA build. It also links respective libraries
from tools/lib and adds tools/include to the include path.
> diff --git a/tools/tracing/rtla/Makefile b/tools/tracing/rtla/Makefile
> index 45690ee14544d..289e44c9664b2 100644
> --- a/tools/tracing/rtla/Makefile
> +++ b/tools/tracing/rtla/Makefile
> @@ -27,6 +27,24 @@ endif
> RTLA := $(OUTPUT)rtla
> RTLA_IN := $(RTLA)-in.o
>
> +LIBSUBCMD_DIR = $(srctree)/tools/lib/subcmd/
> +ifneq ($(OUTPUT),)
> + LIBSUBCMD_OUTPUT = $(abspath $(OUTPUT))/libsubcmd
> +else
> + LIBSUBCMD_OUTPUT = $(CURDIR)/libsubcmd
> +endif
> +LIBSUBCMD = $(LIBSUBCMD_OUTPUT)/libsubcmd.a
> +LIBSUBCMD_INCLUDES = -I$(LIBSUBCMD_OUTPUT)/include
[ ... ]
> -$(RTLA_IN): fixdep FORCE src/timerlat.skel.h
> +$(RTLA_IN): fixdep FORCE src/timerlat.skel.h $(LIBSUBCMD_INCLUDES)
> make $(build)=rtla
>
> -clean: doc_clean fixdep-clean
> +$(LIBSUBCMD_OUTPUT):
> + $(Q)$(MKDIR) -p $@
> +
> +$(LIBSUBCMD_INCLUDES): $(LIBSUBCMD_OUTPUT)
> + $(Q)$(MAKE) -C $(LIBSUBCMD_DIR) $(LIBSUBCMD_MAKEFLAGS) \
> + install_headers
Does this unintentionally cause the install_headers sub-make to run on
every build?
Since LIBSUBCMD_INCLUDES expands to the compiler flag string
-I$(LIBSUBCMD_OUTPUT)/include, make interprets this literal string as a
target. Because a file named "-I..." is never actually created by the
install_headers step, make considers the target permanently out-of-date
and executes this recipe unconditionally on every invocation.
> +$(LIBSUBCMD): fixdep $(LIBSUBCMD_OUTPUT)
> + $(Q)$(MAKE) -C $(LIBSUBCMD_DIR) $(LIBSUBCMD_MAKEFLAGS) \
> + $@
> +
> +$(LIB_STR_ERROR_R): $(LIB_STR_ERROR_R_SRC)
> + $(QUIET_CC)$(CC) $(CFLAGS) -c -o $@ $<
> +
> +$(LIB_STRING): $(LIB_STRING_SRC)
> + $(QUIET_CC)$(CC) $(CFLAGS) -c -o $@ $<
Do these manual compilation rules bypass dependency tracking?
These targets invoke the compiler directly without using the standard
tools/build/Makefile.build infrastructure. By omitting the necessary
compiler flags to generate dependency files and skipping the fixdep
post-processing, updates to any headers included by string.c or
str_error_r.c will not automatically trigger a rebuild of these object
files.
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260320150651.51057-1-tglozar@redhat.com?part=1
next prev parent reply other threads:[~2026-04-25 1:37 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-20 15:06 [PATCH 0/3] rtla: Migrate to libsubcmd for command line option parsing Tomas Glozar
2026-03-20 15:06 ` [PATCH 1/3] rtla: Add libsubcmd dependency Tomas Glozar
2026-04-25 1:37 ` sashiko-bot [this message]
2026-03-20 15:06 ` [PATCH 2/3] tools subcmd: support optarg as separate argument Tomas Glozar
2026-03-20 15:06 ` [PATCH 3/3] rtla: Parse cmdline using libsubcmd Tomas Glozar
2026-03-20 17:31 ` Wander Lairson Costa
2026-03-23 14:15 ` Tomas Glozar
2026-03-21 16:08 ` Costa Shulyupin
2026-03-23 14:26 ` Tomas Glozar
2026-03-24 14:37 ` Tomas Glozar
2026-04-25 2:06 ` sashiko-bot
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=20260425013739.BF4C4C19425@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=sashiko@lists.linux.dev \
--cc=tglozar@redhat.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