Linux-ARM-Kernel Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v2 0/2] arm64: errata: Fix two Kconfig column entries
@ 2026-11-21  8:39 TAKANO Mitsuhiro
  2026-11-21  8:39 ` [PATCH v2 1/2] arm64: errata: Set GIC-700 #2941627 Kconfig entry to N/A TAKANO Mitsuhiro
  2026-11-21  8:39 ` [PATCH v2 2/2] arm64: errata: Fix Falkor E1041 Kconfig name TAKANO Mitsuhiro
  0 siblings, 2 replies; 3+ messages in thread
From: TAKANO Mitsuhiro @ 2026-11-21  8:39 UTC (permalink / raw)
  To: Catalin Marinas, Will Deacon
  Cc: Mark Rutland, Marc Zyngier, Jonathan Corbet, linux-arm-kernel,
	Lorenzo Pieralisi, linux-doc, linux-kernel, TAKANO Mitsuhiro

Documentation/arch/arm64/silicon-errata.rst names two Kconfig symbols that
no Kconfig file defines.  I checked every Kconfig name in the table: these
are the only two.

The GIC-700 #2941627 entry names ARM64_ERRATUM_2941627; that workaround is
a GIC driver quirk, matched at runtime against the GIC's IIDR and enabled
through a static branch, so there is no symbol to name.  The Falkor E1041
entry names QCOM_FALKOR_ERRATUM_1041; the symbol defined in the tree is
QCOM_FALKOR_ERRATUM_E1041.

The two rows come from different commits, so they are kept separate and
each carries its own Fixes: tag.  Both are documentation-only changes: the
table keeps its column widths and no build was run.  I did not add a Cc:
stable tag, as nothing outside Documentation changes.

v2: the change is the same as in v1; only the index and hunk line numbers
moved with the rebase.  One changelog sentence is replaced in each patch:
in 1/2 v1 called the table row the tree's only record of the erratum, but
irq-gic-v3.c records it in gic_quirks[]; in 2/2 the sentence about the
symbol is reworded to name arch/arm64/Kconfig.  Both changelogs now say
that no build was run.

v1 was posted on 6 November 2026:
https://lore.kernel.org/all/20261106234541.1474324-1-takano32@gmail.com/
A duplicate submission went out the same night with different subjects on
1/2 and 2/2 (20261106235616.1533003); please ignore it.

The changes and the changelog text were produced with an AI coding
assistant (Hermes Agent, model deepseek-v4.1-flash) working from the tree
and the git history.  I reviewed the diff and the reasoning, and I am
responsible for the submission.

TAKANO Mitsuhiro (2):
  arm64: errata: Set GIC-700 #2941627 Kconfig entry to N/A
  arm64: errata: Fix Falkor E1041 Kconfig name

 Documentation/arch/arm64/silicon-errata.rst | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)


base-commit: d0f871bfb831705d9f55479f5b2b40f636ea7b16
-- 
2.53.0



^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH v2 1/2] arm64: errata: Set GIC-700 #2941627 Kconfig entry to N/A
  2026-11-21  8:39 [PATCH v2 0/2] arm64: errata: Fix two Kconfig column entries TAKANO Mitsuhiro
@ 2026-11-21  8:39 ` TAKANO Mitsuhiro
  2026-11-21  8:39 ` [PATCH v2 2/2] arm64: errata: Fix Falkor E1041 Kconfig name TAKANO Mitsuhiro
  1 sibling, 0 replies; 3+ messages in thread
From: TAKANO Mitsuhiro @ 2026-11-21  8:39 UTC (permalink / raw)
  To: Catalin Marinas, Will Deacon
  Cc: Mark Rutland, Marc Zyngier, Jonathan Corbet, linux-arm-kernel,
	Lorenzo Pieralisi, linux-doc, linux-kernel, TAKANO Mitsuhiro

The GIC-700 #2941627 entry in the arm64 silicon-errata table names
ARM64_ERRATUM_2941627 in its Kconfig column, but no Kconfig file defines
that symbol and no code refers to it.

The workaround has no Kconfig option: drivers/irqchip/irq-gic-v3.c
passes the GICD_IIDR value to gic_enable_quirks(), which compares it
against gic_quirks[]; the matching quirk,
gic_enable_quirk_arm64_2941627(), enables a static branch.  The table
already records GIC-driver workarounds that have no Kconfig symbol as
"N/A", for example the NVIDIA T241-FABRIC-4 row, whose workaround is
likewise in the GIC driver.

Set the Kconfig column to "N/A" for this row.  Leave the row in place:
only the Kconfig column is wrong, and the table should keep recording the
erratum.

No functional change; this only touches Documentation, so no build was
run.  The change and this changelog were written with the help of an LLM
assistant (Hermes Agent); I reviewed both and am responsible for them.

Fixes: 6fe5c68ee6a1 ("irqchip/gic-v3: Workaround for GIC-700 erratum 2941627")
Signed-off-by: TAKANO Mitsuhiro <takano32@gmail.com>
---
v2: the change is the same as in v1; only the index and hunk line
numbers moved with the rebase.  One changelog sentence is replaced in
each patch: in 1/2 v1 called the table row the tree's only record of
the erratum, but irq-gic-v3.c records it in gic_quirks[]; in 2/2 the
sentence about the symbol is reworded to name arch/arm64/Kconfig.
Both changelogs now say that no build was run.
v1: https://lore.kernel.org/all/20261106234541.1474324-1-takano32@gmail.com/
A duplicate submission went out the same night with different subjects
on 1/2 and 2/2 (20261106235616.1533003); please ignore it.

 Documentation/arch/arm64/silicon-errata.rst | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/arch/arm64/silicon-errata.rst b/Documentation/arch/arm64/silicon-errata.rst
index bffaa46c664b..4bb5707a880a 100644
--- a/Documentation/arch/arm64/silicon-errata.rst
+++ b/Documentation/arch/arm64/silicon-errata.rst
@@ -297,7 +297,7 @@ stable kernels.
 | ARM            | MMU S3          | #3995052,       | N/A                         |
 |                |                 | #3673557        |                             |
 +----------------+-----------------+-----------------+-----------------------------+
-| ARM            | GIC-700         | #2941627        | ARM64_ERRATUM_2941627       |
+| ARM            | GIC-700         | #2941627        | N/A                         |
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | SI L1           | #4311569        | ARM64_ERRATUM_4311569       |
 +----------------+-----------------+-----------------+-----------------------------+
-- 
2.53.0



^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [PATCH v2 2/2] arm64: errata: Fix Falkor E1041 Kconfig name
  2026-11-21  8:39 [PATCH v2 0/2] arm64: errata: Fix two Kconfig column entries TAKANO Mitsuhiro
  2026-11-21  8:39 ` [PATCH v2 1/2] arm64: errata: Set GIC-700 #2941627 Kconfig entry to N/A TAKANO Mitsuhiro
@ 2026-11-21  8:39 ` TAKANO Mitsuhiro
  1 sibling, 0 replies; 3+ messages in thread
From: TAKANO Mitsuhiro @ 2026-11-21  8:39 UTC (permalink / raw)
  To: Catalin Marinas, Will Deacon
  Cc: Mark Rutland, Marc Zyngier, Jonathan Corbet, linux-arm-kernel,
	Lorenzo Pieralisi, linux-doc, linux-kernel, TAKANO Mitsuhiro

The Falkor E1041 entry in the arm64 silicon-errata table names
QCOM_FALKOR_ERRATUM_1041, but the symbol is QCOM_FALKOR_ERRATUM_E1041: it
is defined in arch/arm64/Kconfig and used in
arch/arm64/include/asm/assembler.h.  Correct the name in the table.

The E1003 and E1009 rows name QCOM_FALKOR_ERRATUM_1003 and
QCOM_FALKOR_ERRATUM_1009; the E in the erratum ID is not part of either
symbol name, so an E-less name in the E1041 row looked consistent.
QCOM_FALKOR_ERRATUM_E1041 is the only Falkor symbol that carries the E.

The symbol in arch/arm64/Kconfig is spelled correctly; only the table
entry is wrong.

No functional change; this only touches Documentation, so no build was
run.  The change and this changelog were written with the help of an LLM
assistant (Hermes Agent); I reviewed both and am responsible for them.

Fixes: 932b50c7c1c6 ("arm64: Add software workaround for Falkor erratum 1041")
Signed-off-by: TAKANO Mitsuhiro <takano32@gmail.com>
---
v2: the change is the same as in v1; only the index and hunk line
numbers moved with the rebase.  One changelog sentence is replaced in
each patch: in 1/2 v1 called the table row the tree's only record of
the erratum, but irq-gic-v3.c records it in gic_quirks[]; in 2/2 the
sentence about the symbol is reworded to name arch/arm64/Kconfig.
Both changelogs now say that no build was run.
v1: https://lore.kernel.org/all/20261106234541.1474324-1-takano32@gmail.com/
A duplicate submission went out the same night with different subjects
on 1/2 and 2/2 (20261106235616.1533003); please ignore it.

 Documentation/arch/arm64/silicon-errata.rst | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/arch/arm64/silicon-errata.rst b/Documentation/arch/arm64/silicon-errata.rst
index 4bb5707a880a..cc7bf2b6a09b 100644
--- a/Documentation/arch/arm64/silicon-errata.rst
+++ b/Documentation/arch/arm64/silicon-errata.rst
@@ -378,7 +378,7 @@ stable kernels.
 +----------------+-----------------+-----------------+-----------------------------+
 | Qualcomm Tech. | QDF2400 ITS     | E0065           | QCOM_QDF2400_ERRATUM_0065   |
 +----------------+-----------------+-----------------+-----------------------------+
-| Qualcomm Tech. | Falkor v{1,2}   | E1041           | QCOM_FALKOR_ERRATUM_1041    |
+| Qualcomm Tech. | Falkor v{1,2}   | E1041           | QCOM_FALKOR_ERRATUM_E1041   |
 +----------------+-----------------+-----------------+-----------------------------+
 | Qualcomm Tech. | Kryo4xx Gold    | N/A             | ARM64_ERRATUM_1463225       |
 +----------------+-----------------+-----------------+-----------------------------+
-- 
2.53.0



^ permalink raw reply related	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2026-11-21  8:40 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-21  8:39 [PATCH v2 0/2] arm64: errata: Fix two Kconfig column entries TAKANO Mitsuhiro
2026-11-21  8:39 ` [PATCH v2 1/2] arm64: errata: Set GIC-700 #2941627 Kconfig entry to N/A TAKANO Mitsuhiro
2026-11-21  8:39 ` [PATCH v2 2/2] arm64: errata: Fix Falkor E1041 Kconfig name TAKANO Mitsuhiro

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox