From: Jim Cromie via B4 Relay <devnull+jim.cromie.gmail.com@kernel.org>
To: Arnd Bergmann <arnd@arndb.de>, Jason Baron <jbaron@akamai.com>,
Luis Chamberlain <mcgrof@kernel.org>,
Petr Pavlu <petr.pavlu@suse.com>,
Daniel Gomez <da.gomez@kernel.org>,
Sami Tolvanen <samitolvanen@google.com>,
Aaron Tomlin <atomlin@atomlin.com>,
Andrew Morton <akpm@linux-foundation.org>
Cc: linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org,
linux-modules@vger.kernel.org, Jim Cromie <jim.cromie@gmail.com>,
Louis Chauvet <louis.chauvet@bootlin.com>
Subject: [PATCH 3/6] vmlinux.lds.h: Fix ALIGN(8) omission causing NULL ptr on i386
Date: Sat, 21 Nov 2026 15:49:21 -0700 [thread overview]
Message-ID: <20261121-vmlinux-lds-v1-3-49818f3f5cc0@gmail.com> (raw)
In-Reply-To: <20261121-vmlinux-lds-v1-0-49818f3f5cc0@gmail.com>
From: Jim Cromie <jim.cromie@gmail.com>
Almost all uses of the BOUNDED_SECTION macros are ALIGN(8), either
explicitly, or by being below an aligned section containing x*8 byte
objects. The noteworthy exception is BOUNDED_SECTION(__dyndbg), which
immediately follows BOUNDED_SECTION(__dyndbg_classes).
On i386, struct _ddebug_classmap is 28 bytes, so without an explicit
ALIGN(8) in the macro, the following __dyndbg section gets misaligned,
causing a NULL ptr deref in dynamic_debug_init().
So fix this with an explicit ALIGN(8) in the existing BOUNDED_SECTION
macros, and introduce _ALIGNED variants to handle the cases with an
explicit . = ALIGN(x)
Also add explicit alignments for: EXCEPTION_TABLE, ORC_UNWIND_TABLE,
TRACEDATA, INIT_SETUP, and NOTES.
update BOUNDED_SECTION uses inside . = ALIGN(x) stanzas to use
_ALIGNED variants, but keep the outer ALIGNs so the symbols between
them are not "re-aligned".
In particular, scripts/sorttable.c does not tolerate sloppy padding.
At the top of ORC_UNWIND_TABLE, add . = ALIGN(4) to match the struct
orc_header __align() call in the code:
commit b9f174c811e3 ("x86/unwind/orc: Add ELF section with ORC version identifier")
Suggested-by: Louis Chauvet <louis.chauvet@bootlin.com> # _ALIGNED variants.
Link: https://lore.kernel.org/lkml/177402491426.6181.12855763650074831089.b4-review@b4/
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
---
v3:
sashiko complained about NOTES and .BTF_ids.
gemini asserts that NOTES are natively 4-byte aligned, add comment repeating it.
.BTF_ids doesnt use BOUNDED_BY, since start/end isnt needed;
sashiko evidently got confused by immediately preceding usage.
v2:
sashiko picked up 2 cases, added to the explicit list above
https://sashiko.dev/#/patchset/20260515-asm-generic-1-v3-0-680b273666d4%40gmail.com
---
include/asm-generic/bounded_sections.lds.h | 17 ++++++++++++++---
include/asm-generic/vmlinux.lds.h | 18 ++++++++++--------
2 files changed, 24 insertions(+), 11 deletions(-)
diff --git a/include/asm-generic/bounded_sections.lds.h b/include/asm-generic/bounded_sections.lds.h
index 268cdc34389b..8ff3e3420f60 100644
--- a/include/asm-generic/bounded_sections.lds.h
+++ b/include/asm-generic/bounded_sections.lds.h
@@ -3,19 +3,30 @@
#ifndef _ASM_GENERIC_BOUNDED_SECTIONS_H
#define _ASM_GENERIC_BOUNDED_SECTIONS_H
-#define BOUNDED_SECTION_PRE_LABEL(_sec_, _label_, _BEGIN_, _END_) \
+#define BOUNDED_SECTION_PRE_LABEL_ALIGNED(_sec_, _label_, _BEGIN_, _END_, _ALIGNED_) \
+ . = ALIGN(_ALIGNED_); \
_BEGIN_##_label_ = .; \
KEEP(*(_sec_)) \
_END_##_label_ = .;
-#define BOUNDED_SECTION_POST_LABEL(_sec_, _label_, _BEGIN_, _END_) \
+#define BOUNDED_SECTION_PRE_LABEL(_sec_, _label_, _BEGIN_, _END_) \
+ BOUNDED_SECTION_PRE_LABEL_ALIGNED(_sec_, _label_, _BEGIN_, _END_, 8)
+
+#define BOUNDED_SECTION_POST_LABEL_ALIGNED(_sec_, _label_, _BEGIN_, _END_, _ALIGNED_) \
+ . = ALIGN(_ALIGNED_); \
_label_##_BEGIN_ = .; \
KEEP(*(_sec_)) \
_label_##_END_ = .;
+#define BOUNDED_SECTION_POST_LABEL(_sec_, _label_, _BEGIN_, _END_) \
+ BOUNDED_SECTION_POST_LABEL_ALIGNED(_sec_, _label_, _BEGIN_, _END_, 8)
+
#define BOUNDED_SECTION_BY(_sec_, _label_) \
BOUNDED_SECTION_PRE_LABEL(_sec_, _label_, __start, __stop)
-#define BOUNDED_SECTION(_sec) BOUNDED_SECTION_BY(_sec, _sec)
+#define BOUNDED_SECTION_BY_ALIGNED(_sec_, _label_, _ALIGNED_) \
+ BOUNDED_SECTION_PRE_LABEL_ALIGNED(_sec_, _label_, __start, __stop, _ALIGNED_)
+
+#define BOUNDED_SECTION(_sec) BOUNDED_SECTION_BY(_sec, _sec)
#endif /* _ASM_GENERIC_BOUNDED_SECTIONS_H */
diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index 5fed4194983c..c570495d8de5 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -640,7 +640,7 @@
#define EXCEPTION_TABLE(align) \
. = ALIGN(align); \
__ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(__ex_table, ___ex_table) \
+ BOUNDED_SECTION_BY_ALIGNED(__ex_table, ___ex_table, align) \
}
/*
@@ -679,7 +679,7 @@
#define BTF \
. = ALIGN(PAGE_SIZE); \
.BTF : AT(ADDR(.BTF) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(.BTF, _BTF) \
+ BOUNDED_SECTION_BY_ALIGNED(.BTF, _BTF, PAGE_SIZE) \
} \
BTF_INLINE \
BTF_INLINE_LINK \
@@ -871,16 +871,17 @@
#ifdef CONFIG_UNWINDER_ORC
#define ORC_UNWIND_TABLE \
+ . = ALIGN(4); \
.orc_header : AT(ADDR(.orc_header) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(.orc_header, _orc_header) \
+ BOUNDED_SECTION_BY_ALIGNED(.orc_header, _orc_header, 4) \
} \
. = ALIGN(4); \
.orc_unwind_ip : AT(ADDR(.orc_unwind_ip) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(.orc_unwind_ip, _orc_unwind_ip) \
+ BOUNDED_SECTION_BY_ALIGNED(.orc_unwind_ip, _orc_unwind_ip, 4)\
} \
. = ALIGN(2); \
.orc_unwind : AT(ADDR(.orc_unwind) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(.orc_unwind, _orc_unwind) \
+ BOUNDED_SECTION_BY_ALIGNED(.orc_unwind, _orc_unwind, 2) \
} \
text_size = _etext - _stext; \
. = ALIGN(4); \
@@ -908,7 +909,7 @@
#define TRACEDATA \
. = ALIGN(4); \
.tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { \
- BOUNDED_SECTION_POST_LABEL(.tracedata, __tracedata, _start, _end) \
+ BOUNDED_SECTION_POST_LABEL_ALIGNED(.tracedata, __tracedata, _start, _end, 4) \
}
#else
#define TRACEDATA
@@ -937,13 +938,14 @@
*(.note.gnu.property) \
} \
.notes : AT(ADDR(.notes) - LOAD_OFFSET) { \
- BOUNDED_SECTION_BY(.note.*, _notes) \
+ /* *(.note.*) are natively 4-byte aligned */ \
+ BOUNDED_SECTION_BY_ALIGNED(.note.*, _notes, 4) \
} NOTES_HEADERS \
NOTES_HEADERS_RESTORE
#define INIT_SETUP(initsetup_align) \
. = ALIGN(initsetup_align); \
- BOUNDED_SECTION_POST_LABEL(.init.setup, __setup, _start, _end)
+ BOUNDED_SECTION_POST_LABEL_ALIGNED(.init.setup, __setup, _start, _end, initsetup_align)
#define INIT_CALLS_LEVEL(level) \
__initcall##level##_start = .; \
--
2.56.1
next prev parent reply other threads:[~2026-11-21 22:49 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-21 22:49 [PATCH 0/6] dyndbg,vmlinux.lds.h: fix i386 section alignment in BOUNDED_SECTION macros Jim Cromie via B4 Relay
2026-11-21 22:49 ` [PATCH 1/6] vmlinux.lds.h: refactor BOUNDED_SECTION_* macros into bounded_sections.lds.h Jim Cromie via B4 Relay
2026-11-21 22:49 ` [PATCH 2/6] vmlinux.lds.h: drop unused HEADERED_SECTION* macros Jim Cromie via B4 Relay
2026-11-21 22:49 ` Jim Cromie via B4 Relay [this message]
2026-11-21 22:49 ` [PATCH 4/6] vmlinux.lds.h: remove redundant ALIGN(8) directives Jim Cromie via B4 Relay
2026-11-21 22:49 ` [PATCH 5/6] dyndbg.lds.S: fix lost dyndbg sections in modules Jim Cromie via B4 Relay
2026-11-21 22:49 ` [PATCH 6/6] checkpatch.pl: generalize linker script macro exceptions Jim Cromie via B4 Relay
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=20261121-vmlinux-lds-v1-3-49818f3f5cc0@gmail.com \
--to=devnull+jim.cromie.gmail.com@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=atomlin@atomlin.com \
--cc=da.gomez@kernel.org \
--cc=jbaron@akamai.com \
--cc=jim.cromie@gmail.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-modules@vger.kernel.org \
--cc=louis.chauvet@bootlin.com \
--cc=mcgrof@kernel.org \
--cc=petr.pavlu@suse.com \
--cc=samitolvanen@google.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
all inboxes | Powered by JetHome®