From: Masahiro Yamada <masahiroy@kernel.org>
To: linux-kbuild@vger.kernel.org
Cc: linux-kernel@vger.kernel.org,
Andrea della Porta <andrea.porta@suse.com>,
linux-clk@vger.kernel.org, Stephen Boyd <sboyd@kernel.org>,
Brendan Higgins <brendanhiggins@google.com>,
Rob Herring <robh@kernel.org>,
devicetree@vger.kernel.org,
Masahiro Yamada <masahiroy@kernel.org>,
Nathan Chancellor <nathan@kernel.org>,
Nicolas Schier <nicolas@fjasle.eu>
Subject: [PATCH v3 1/2] kbuild: move non-boot built-in DTBs to .rodata section
Date: Mon, 23 Sep 2024 16:56:02 +0900 [thread overview]
Message-ID: <20240923075704.3567313-1-masahiroy@kernel.org> (raw)
Commit aab94339cd85 ("of: Add support for linking device tree blobs
into vmlinux") introduced a mechanism to embed DTBs into vmlinux.
Initially, it was used for wrapping boot DTBs in arch/*/boot/dts/, but
it is now reused for more generic purposes, such as testing.
Built-in DTBs are discarded because KERNEL_DTB() is part of INIT_DATA,
as defined in include/asm-generic/vmlinux.lds.h.
This has not been an issue so far because OF unittests are triggered
during boot, as defined by late_initcall(of_unittest).
However, the recent clk KUnit test additions have caused problems
because KUnit can execute test suites after boot.
For example:
# echo > /sys/kernel/debug/kunit/clk_register_clk_parent_data_device/run
This command triggers a stack trace because built-in DTBs have already
been freed.
While it is possible to move such test suites from kunit_test_suites to
kunit_test_init_section_suites, it would be preferable to avoid usage
limitations.
This commit moves non-boot built-in DTBs to the .rodata section. Since
these generic DTBs are looked up by name, they do not need to be placed
in the special .dtb.init.rodata section.
Boot DTBs should remain in .dtb.init.rodata because the arch boot code
generally does not know the DT name, thus it uses the __dtb_start symbol
to locate it.
This separation also ensures that the __dtb_start symbol references the
boot DTB. Currently, the .dtb.init.rodata is a mixture of both boot and
non-boot DTBs. The __dtb_start symbol must be followed by the boot DTB,
but we currently rely on the link order (i.e., the order in Makefiles),
which is very fragile.
Fixes: 5c9dd72d8385 ("of: Add a KUnit test for overlays and test managed APIs")
Fixes: 5776526beb95 ("clk: Add KUnit tests for clk fixed rate basic type")
Fixes: 274aff8711b2 ("clk: Add KUnit tests for clks registered with struct clk_parent_data")
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
---
Changes in v3:
- Move to .rodata section instead of .init.rodata
scripts/Makefile.dtbs | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/scripts/Makefile.dtbs b/scripts/Makefile.dtbs
index 46009d5f1486..3e81cca6f68a 100644
--- a/scripts/Makefile.dtbs
+++ b/scripts/Makefile.dtbs
@@ -34,12 +34,14 @@ $(obj)/dtbs-list: $(dtb-y) FORCE
# Assembly file to wrap dtb(o)
# ---------------------------------------------------------------------------
+builtin-dtb-section = $(if $(filter arch/%, $(obj)),.dtb.init.rodata,.rodata)
+
# Generate an assembly file to wrap the output of the device tree compiler
quiet_cmd_wrap_S_dtb = WRAP $@
cmd_wrap_S_dtb = { \
symbase=__$(patsubst .%,%,$(suffix $<))_$(subst -,_,$(notdir $*)); \
echo '\#include <asm-generic/vmlinux.lds.h>'; \
- echo '.section .dtb.init.rodata,"a"'; \
+ echo '.section $(builtin-dtb-section),"a"'; \
echo '.balign STRUCT_ALIGNMENT'; \
echo ".global $${symbase}_begin"; \
echo "$${symbase}_begin:"; \
--
2.43.0
next reply other threads:[~2024-09-23 7:57 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-09-23 7:56 Masahiro Yamada [this message]
2024-09-23 7:56 ` [PATCH v3 2/2] kbuild: add generic support for built-in boot DTBs Masahiro Yamada
2024-09-24 22:15 ` [PATCH v3 1/2] kbuild: move non-boot built-in DTBs to .rodata section Rob Herring
2024-09-27 14:51 ` Masahiro Yamada
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=20240923075704.3567313-1-masahiroy@kernel.org \
--to=masahiroy@kernel.org \
--cc=andrea.porta@suse.com \
--cc=brendanhiggins@google.com \
--cc=devicetree@vger.kernel.org \
--cc=linux-clk@vger.kernel.org \
--cc=linux-kbuild@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=nathan@kernel.org \
--cc=nicolas@fjasle.eu \
--cc=robh@kernel.org \
--cc=sboyd@kernel.org \
/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