bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Yonghong Song <yonghong.song@linux.dev>
To: Paul Menzel <pmenzel@molgen.mpg.de>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Andrii Nakryiko <andrii@kernel.org>
Cc: bpf@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: How to deal with increased install size with `CONFIG_DEBUG_INFO_BTF=y`
Date: Mon, 25 Aug 2025 10:54:04 -0700	[thread overview]
Message-ID: <a2d8fba6-288c-4d2c-ac2c-d7a920c0cf39@linux.dev> (raw)
In-Reply-To: <15bcb208-3ccb-4d99-853b-545626914373@molgen.mpg.de>



On 8/25/25 8:42 AM, Paul Menzel wrote:
> Dear Linux folks,
>
>
> Trying to get ptcpdump [1] running, I needed to build Linux with BTF 
> symbols.
>
> ```
> @@ -5985,11 +5986,22 @@
>  #
>  # Compile-time checks and compiler options
>  #
> +CONFIG_DEBUG_INFO=y
>  CONFIG_AS_HAS_NON_CONST_ULEB128=y
> -CONFIG_DEBUG_INFO_NONE=y
> +# CONFIG_DEBUG_INFO_NONE is not set
>  # CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT is not set
>  # CONFIG_DEBUG_INFO_DWARF4 is not set
> -# CONFIG_DEBUG_INFO_DWARF5 is not set
> +CONFIG_DEBUG_INFO_DWARF5=y
> +# CONFIG_DEBUG_INFO_REDUCED is not set
> +CONFIG_DEBUG_INFO_COMPRESSED_NONE=y
> +# CONFIG_DEBUG_INFO_COMPRESSED_ZLIB is not set
> +# CONFIG_DEBUG_INFO_SPLIT is not set
> +CONFIG_DEBUG_INFO_BTF=y
> +CONFIG_PAHOLE_HAS_SPLIT_BTF=y
> +CONFIG_PAHOLE_HAS_LANG_EXCLUDE=y
> +CONFIG_DEBUG_INFO_BTF_MODULES=y
> +# CONFIG_MODULE_ALLOW_BTF_MISMATCH is not set
> +# CONFIG_GDB_SCRIPTS is not set
>  CONFIG_FRAME_WARN=2048
>  # CONFIG_STRIP_ASM_SYMS is not set
>  # CONFIG_READABLE_ASM is not set
> ```
>
> This increased the module size quite a bit:
>
>     $ du -sh /lib/modules/6.12.4*
>     128M        /lib/modules/6.12.40.mx64.484
>     1.9G        /lib/modules/6.12.43.mx64.485

The increase is mostly due to dwarf. The BTF related sections should be just ~6M.

>
> Searching the WWW, it was suggested to run `INSTALL_MOD_STRIP=1` [2], 
> bringing it down to 137 MB:
>
>     137M        /lib/modules/6.12.43.mx64.486
>
> I was under the expression, that BTF symbols are small compared to the 
> debug symbol types from the past. (Excuse my ignorance about the 
> terminology and subject.) Is `INSTALL_MOD_STRIP=1` the “recommended” 
> way to still use BPF/BTF on production systems.

You can use INSTALL_MOD_STRIP=1 to reduce kernel binary size in /lib/modules/... directory.

>
>
> Kind regards,
>
> Paul
>
>
> [1]: https://github.com/mozillazg/ptcpdump
> [2]: 
> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/28218/diffs
>


      reply	other threads:[~2025-08-25 17:54 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-25 15:42 How to deal with increased install size with `CONFIG_DEBUG_INFO_BTF=y` Paul Menzel
2025-08-25 17:54 ` Yonghong Song [this message]

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=a2d8fba6-288c-4d2c-ac2c-d7a920c0cf39@linux.dev \
    --to=yonghong.song@linux.dev \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pmenzel@molgen.mpg.de \
    /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).