linux-kernel - Re: [PATCH] tools/build: Skip jobserver flgas in -s detection

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAK7LNARKnzM8hX1P1Gt8MDb6DxO4Nys7k9vphwqsrVW29vZg8g@mail.gmail.com>
Date: Sat, 8 Feb 2025 06:14:10 +0900
From: Masahiro Yamada <masahiroy@...nel.org>
To: Daniel Xu <dxu@...uu.xyz>
Cc: linux-kbuild@...r.kernel.org, linux-kernel@...r.kernel.org, 
	bpf@...r.kernel.org, jolsa@...nel.org
Subject: Re: [PATCH] tools/build: Skip jobserver flgas in -s detection

On Sun, Feb 2, 2025 at 5:11 AM Daniel Xu <dxu@...uu.xyz> wrote:
>
> Currently there is unnecessarily verbose output:
>
>     $ make -j8 bzImage
>     mkdir -p /home/dlxu/dev/linux/tools/objtool && make
>       O=/home/dlxu/dev/linux subdir=tools/objtool --no-print-directory -C
>       objtool
>     mkdir -p /home/dlxu/dev/linux/tools/bpf/resolve_btfids && make
>       O=/home/dlxu/dev/linux subdir=tools/bpf/resolve_btfids
>       --no-print-directory -C bpf/resolve_btfids
>       INSTALL libsubcmd_headers
>       INSTALL libsubcmd_headers
>       UPD     include/config/kernel.release
>
> The reason this happens is that it seems that make is internally adding
> the following flag to $(MAKEFLAGS):
>
>     ---jobserver-auth=fifo:/tmp/GMfifo1880691
>
> This breaks -s detection which searches for 's' in $(short-opts), as any
> this entire long flag is treated as a short flag and the presence of any
> 's' triggers silent=1.
>
> Fix by filtering out such flags so it's still correct to do a substring
> search for 's'.
>
> Signed-off-by: Daniel Xu <dxu@...uu.xyz>


This is not the right fix.

The code for calculating short-opts is correct.

This is documented in GNU Make manual:
 Recall that MAKEFLAGS will put all single-letter options (such as ‘-t’) into
 the first word, and that word will be empty if no single-letter options were
 given. To work with this, it’s helpful to add a value at the start to ensure
 there’s a word: for example ‘-$(MAKEFLAGS)’.

https://www.gnu.org/software/make/manual/make.html#Testing-Flags


The root cause is different.




> ---
>  tools/scripts/Makefile.include | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tools/scripts/Makefile.include b/tools/scripts/Makefile.include
> index 0aa4005017c7..a413f73a7856 100644
> --- a/tools/scripts/Makefile.include
> +++ b/tools/scripts/Makefile.include
> @@ -139,9 +139,9 @@ endif
>  # If the user is running make -s (silent mode), suppress echoing of commands
>  # make-4.0 (and later) keep single letter options in the 1st word of MAKEFLAGS.
>  ifeq ($(filter 3.%,$(MAKE_VERSION)),)
> -short-opts := $(firstword -$(MAKEFLAGS))
> +short-opts := $(filter-out ---%,$(firstword -$(MAKEFLAGS)))
>  else
> -short-opts := $(filter-out --%,$(MAKEFLAGS))
> +short-opts := $(filter-out --% ---%,$(MAKEFLAGS))
>  endif
>
>  ifneq ($(findstring s,$(short-opts)),)
> --
> 2.47.1
>
>


--
Best Regards
Masahiro Yamada