Linux Modules
 help / color / mirror / Atom feed
From: Jim Cromie <jim.cromie@gmail.com>
To: Andrew Morton <akpm@linux-foundation.org>,
	 Jason Baron <jbaron@akamai.com>,
	 Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	 Shuah Khan <skhan@linuxfoundation.org>,
	Simona Vetter <simona@ffwll.ch>,
	 Luis Chamberlain <mcgrof@kernel.org>,
	Petr Pavlu <petr.pavlu@suse.com>
Cc: linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
	 dri-devel@lists.freedesktop.org, linux-arch@vger.kernel.org,
	 linux-modules@vger.kernel.org, linux-kselftest@vger.kernel.org,
	 Jim Cromie <jim.cromie@gmail.com>, Arnd Bergmann <arnd@arndb.de>
Subject: [PATCH v7 10/29] dyndbg.lds.S: fix lost dyndbg sections in modules
Date: Tue, 21 Jul 2026 14:56:59 -0600	[thread overview]
Message-ID: <20260721-dd-maint-2-v7-10-010fbe73b311@gmail.com> (raw)
In-Reply-To: <20260721-dd-maint-2-v7-0-010fbe73b311@gmail.com>

With CONFIG_DRM_USE_DYNAMIC_DEBUG=y, several build configs had
problems with __dyndbg* sections getting lost in drm drivers.  Fix
this by following the model demonstrated in codetag.lds.h.

Introduce include/asm-generic/dyndbg.lds.h, to bundle dynamic-debug's
multiple sections together, into 2 macros:

vmlinux.lds.h DATA_DATA: move the 2 BOUNDED_SECTION_BY(__dyndbg*)
calls into dyndbg.lds.h DYNDBG_SECTIONS(). vmlinux.lds.h now includes
the new file and calls the new macro.

MOD_DYNDBG_SECTIONS keeps the 2 sections by name, aligns them and sets
the output address to 0 when the sections are empty.

dyndbg.lds.h includes (reuses) bounded-section.lds.h

scripts/module.lds.S: now calls MOD_DYNDBG_SECTIONS right before the
CODETAG macro (consistent with their placements in vmlinux.lds.h), and
also includes dyndbg.lds.h

This isolates vmlinux.lds.h from further __dyndbg section additions.

CC: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Reviewed-by: Petr Pavlu <petr.pavlu@suse.com>
---
v3: move #includes to top, drop extra ALIGN(8) in DYNDBG_SECTIONS, add RvBy

v2: Address linker script review feedback for relocatable modules.

MOD_DYNDBG_SECTIONS() used the BOUNDED_SECTION_BY() macro, which
proved problematic for kernel modules for two reasons:

1. Unwanted Empty Sections:
   BOUNDED_SECTION_BY() automatically generates `__start` and `__stop`
   symbols. When applied to `MOD_DYNDBG_SECTIONS()`, the linker assumes
   the sections are populated due to the symbol definitions, forcing an
   empty `__dyndbg` and `__dyndbg_classes` output section in every
   compiled module, even those without dynamic debug configuration.
   Since the module loader uses `section_objs()` to locate data via
   ELF headers instead of relying on `__start`/`__stop` symbols, these
   assignments are completely unnecessary.

2. Non-zero Output Addresses:
   During relocatable linking (e.g., `ld.bfd -r`), omitting an explicit
   base address causes the section to inherit the current location
   counter. This results in non-zero sh_addr values in `.ko` files,
   which is confusing, degrades compressibility, and can cause issues
   with external tools parsing the ELF.

Fix both issues by dropping `BOUNDED_SECTION_BY()` in favor of a simple
`KEEP(*(...))` constraint and explicitly defining the sections with a `0`
base address: `__dyndbg 0 : ALIGN(8) { ... }`.

fixup-inc-vml
---
 MAINTAINERS                       |  1 +
 include/asm-generic/dyndbg.lds.h  | 18 ++++++++++++++++++
 include/asm-generic/vmlinux.lds.h |  6 ++----
 scripts/module.lds.S              |  2 ++
 4 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index dd6e351d2a98..ac050552923e 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -9179,6 +9179,7 @@ DYNAMIC DEBUG
 M:	Jason Baron <jbaron@akamai.com>
 M:	Jim Cromie <jim.cromie@gmail.com>
 S:	Maintained
+F:	include/asm-generic/dyndbg.lds.h
 F:	include/linux/dynamic_debug.h
 F:	lib/dynamic_debug.c
 F:	lib/test_dynamic_debug.c
diff --git a/include/asm-generic/dyndbg.lds.h b/include/asm-generic/dyndbg.lds.h
new file mode 100644
index 000000000000..9d8951bef688
--- /dev/null
+++ b/include/asm-generic/dyndbg.lds.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef __ASM_GENERIC_DYNDBG_LDS_H
+#define __ASM_GENERIC_DYNDBG_LDS_H
+
+#include <asm-generic/bounded_sections.lds.h>
+#define DYNDBG_SECTIONS()					\
+	BOUNDED_SECTION_BY(__dyndbg, ___dyndbg)			\
+	BOUNDED_SECTION_BY(__dyndbg_classes, ___dyndbg_classes)
+
+#define MOD_DYNDBG_SECTIONS()						\
+	__dyndbg 0 : ALIGN(8) {						\
+		KEEP(*(__dyndbg))					\
+	}								\
+	__dyndbg_classes 0 : ALIGN(8) {					\
+		KEEP(*(__dyndbg_classes))				\
+	}
+
+#endif /* __ASM_GENERIC_DYNDBG_LDS_H */
diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index 3758a79d0430..bd60f278f762 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -52,6 +52,7 @@
 
 #include <asm-generic/bounded_sections.lds.h>
 #include <asm-generic/codetag.lds.h>
+#include <asm-generic/dyndbg.lds.h>
 
 #ifndef LOAD_OFFSET
 #define LOAD_OFFSET 0
@@ -344,10 +345,7 @@
 	*(.data..do_once)						\
 	STRUCT_ALIGN();							\
 	*(__tracepoints)						\
-	/* implement dynamic printk debug */				\
-	. = ALIGN(8);							\
-	BOUNDED_SECTION_BY(__dyndbg_classes, ___dyndbg_classes)		\
-	BOUNDED_SECTION_BY(__dyndbg, ___dyndbg)				\
+	DYNDBG_SECTIONS()						\
 	CODETAG_SECTIONS()						\
 	LIKELY_PROFILE()		       				\
 	BRANCH_PROFILE()						\
diff --git a/scripts/module.lds.S b/scripts/module.lds.S
index b62683061d79..2e62dc5bd5d4 100644
--- a/scripts/module.lds.S
+++ b/scripts/module.lds.S
@@ -10,6 +10,7 @@
 #endif
 
 #include <asm-generic/codetag.lds.h>
+#include <asm-generic/dyndbg.lds.h>
 
 SECTIONS {
 	/DISCARD/ : {
@@ -61,6 +62,7 @@ SECTIONS {
 		*(.rodata..L*)
 	}
 
+	MOD_DYNDBG_SECTIONS()
 	MOD_SEPARATE_CODETAG_SECTIONS()
 }
 

-- 
2.55.0


  parent reply	other threads:[~2026-07-21 20:57 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-21 20:56 [PATCH v7 00/29] fix dynamic-debug classmaps API for DRM Jim Cromie
2026-07-21 20:56 ` [PATCH v7 01/29] params: fix a pr_debug(" %p ") use - already in MM-* Jim Cromie
2026-07-21 20:56 ` [PATCH v7 02/29] selftests/dyndbg: Add kselftest script to verify dynamic-debug Jim Cromie
2026-07-21 21:06   ` sashiko-bot
2026-07-21 20:56 ` [PATCH v7 03/29] drm: Fix incorrect ccflags-y spelling inside Makefile Jim Cromie
2026-07-21 20:56 ` [PATCH v7 04/29] drm: fix config dependent unused variable warning Jim Cromie
2026-07-21 20:56 ` [PATCH v7 05/29] drm: Mark CONFIG_DRM_USE_DYNAMIC_DEBUG as unBROKEN Jim Cromie
2026-07-21 20:56 ` [PATCH v7 06/29] vmlinux.lds.h: refactor BOUNDED_SECTION_* macros into bounded_sections.lds.h Jim Cromie
2026-07-21 20:56 ` [PATCH v7 07/29] vmlinux.lds.h: drop unused HEADERED_SECTION* macros Jim Cromie
2026-07-21 20:56 ` [PATCH v7 08/29] vmlinux.lds.h: Fix ALIGN(8) omission causing NULL ptr on i386 Jim Cromie
2026-07-21 20:56 ` [PATCH v7 09/29] vmlinux.lds.h: remove redundant ALIGN(8) directives Jim Cromie
2026-07-21 20:56 ` Jim Cromie [this message]
2026-07-21 20:57 ` [PATCH v7 11/29] dyndbg: factor ddebug_match_desc out from ddebug_change Jim Cromie
2026-07-21 21:05   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 12/29] dyndbg: add stub macro for DECLARE_DYNDBG_CLASSMAP Jim Cromie
2026-07-21 20:57 ` [PATCH v7 13/29] dyndbg: reword "class unknown," to "class:_UNKNOWN_" Jim Cromie
2026-07-21 20:57 ` [PATCH v7 14/29] dyndbg-API: remove DD_CLASS_TYPE_(DISJOINT|LEVEL)_NAMES and code Jim Cromie
2026-07-21 20:57 ` [PATCH v7 15/29] dyndbg: drop NUM_TYPE_ARGS Jim Cromie
2026-07-21 20:57 ` [PATCH v7 16/29] dyndbg: bump num-tokens in a query-cmd from 9 to 15 Jim Cromie
2026-07-21 20:57 ` [PATCH v7 17/29] dyndbg: reduce verbose/debug clutter Jim Cromie
2026-07-21 20:57 ` [PATCH v7 18/29] lib/parser: add match_wildcard_hyphen() for agnostic matching Jim Cromie
2026-07-21 20:57 ` [PATCH v7 19/29] dyndbg: use KBUILD_MODFILE for unique builtin module names Jim Cromie
2026-07-21 20:57 ` [PATCH v7 20/29] dyndbg: refactor param_set_dyndbg_classes and below Jim Cromie
2026-07-21 21:11   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 21/29] dyndbg: tighten fn-sig of ddebug_apply_class_bitmap Jim Cromie
2026-07-21 20:57 ` [PATCH v7 22/29] dyndbg: replace classmap list with an array-slice Jim Cromie
2026-07-21 20:57 ` [PATCH v7 23/29] dyndbg: macrofy a 2-index for-loop pattern Jim Cromie
2026-07-21 20:57 ` [PATCH v7 24/29] dyndbg: pin class param storage to u32 Jim Cromie
2026-07-21 21:10   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 25/29] dyndbg,module: make proper substructs in _ddebug_info Jim Cromie
2026-07-21 21:09   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 26/29] dyndbg: move mod_name down from struct ddebug_table to _ddebug_info Jim Cromie
2026-07-21 21:13   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 27/29] dyndbg: hoist classmap-filter-by-modname up to ddebug_add_module Jim Cromie
2026-07-21 20:57 ` [PATCH v7 28/29] dyndbg-API: replace DECLARE_DYNDBG_CLASSMAP Jim Cromie
2026-07-21 21:18   ` sashiko-bot
2026-07-21 20:57 ` [PATCH v7 29/29] selftests/dyndbg: enable FT_classmap_inheritance Jim Cromie
2026-07-21 21:11   ` 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=20260721-dd-maint-2-v7-10-010fbe73b311@gmail.com \
    --to=jim.cromie@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=arnd@arndb.de \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=jbaron@akamai.com \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=linux-modules@vger.kernel.org \
    --cc=mcgrof@kernel.org \
    --cc=petr.pavlu@suse.com \
    --cc=simona@ffwll.ch \
    --cc=skhan@linuxfoundation.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