[PATCH v1 0/6] Drop duplicate macro definitions

[PATCH v1 0/6] Drop duplicate macro definitions

[Aristo Chen]

Six files define one or two macros twice, with the same value and under
the same preprocessor conditions: two SPI NOR opcodes, two SCSI opcodes,
an ACPI PPTT cache type, two GT-64120 register offsets, an i.MX8 power
domain name and two constants in the video test. An identical
redefinition is valid C, so the compiler never warned about them. This
series drops the second copy, one patch per file.

Three of them need a word of explanation. In include/scsi.h the two
SCSI_SEND_DIAG copies also disagreed on the command type, optional in
one table and mandatory in the other. SCSI-2 makes SEND DIAGNOSTIC
mandatory for all device types, so the copy patch 2 keeps now says
mandatory. include/gt64120.h was imported from Linux, which had the
same duplicate until the fix for it was applied to mips-next [1].
Apart from that, Linux has only changed the header's SPDX tag since the
import, so rather than re-syncing, patch 4 drops the duplicate here as
well. In test/dm/video.c both copies sit inside test functions,
although check_vidconsole_output() further down uses SCROLL_LINES too,
so patch 6 defines both constants once at file scope instead.

No functional change is intended. Built for sandbox, malta,
imx8qxp_mek, imx8qm_mek, qemu-arm-sbsa, rpi_arm64 and evb-ast2600,
which between them cover every touched file, all objects rebuilt by
the series are identical before and after once debug info is stripped
and the embedded version string and timestamp are held fixed, and so
are the final images and the i.MX8 device trees. The only exception is
test/dm/video.o, and with it the sandbox binary: its assertions record
__LINE__ values that patch 6 moves by three lines. Each patch builds
for all seven boards, and the sandbox video tests pass.

The patches are independent of each other and touch different areas
(SPI NOR, SCSI, ACPI, the GT-64120 PCI driver, the i.MX8 device tree
bindings and the video test), so they can be applied one by one or as
a whole. The series is based on next.

[1] https://lore.kernel.org/all/20261001053016.76578-1-aristo.chen@canonical.com/

Aristo Chen (6):
  mtd: spi-nor: drop the duplicate EN4B/EX4B opcode definitions
  scsi: drop the duplicate COMPARE and SEND DIAGNOSTIC opcodes
  acpi: drop the duplicate ACPI_PPTT_CACHE_TYPE_DATA definition
  pci: gt64120: drop the duplicate GT_PCI1M1LD/HD_OFS definitions
  dt-bindings: soc: imx8_pd: drop the duplicate PD_LVDS0_PWM definition
  test: video: define WHITE and SCROLL_LINES once at file scope

 include/acpi/acpi_table.h         | 1 -
 include/dt-bindings/soc/imx8_pd.h | 1 -
 include/gt64120.h                 | 2 --
 include/linux/mtd/spi-nor.h       | 2 --
 include/scsi.h                    | 4 +---
 test/dm/video.c                   | 9 +++------
 6 files changed, 4 insertions(+), 15 deletions(-)

-- 
2.43.0

[PATCH v1 1/6] mtd: spi-nor: drop the duplicate EN4B/EX4B opcode definitions

[Aristo Chen]

SPINOR_OP_EN4B and SPINOR_OP_EX4B are defined twice in a row with the
same values. The second pair came in with the Macronix Octal DTR
opcodes in commit df3d5f9e410c ("mtd: spi-nor: add support for
Macronix Octal flash").

Drop the second pair. No functional change.

Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 include/linux/mtd/spi-nor.h | 2 --
 1 file changed, 2 deletions(-)

diff --git a/include/linux/mtd/spi-nor.h b/include/linux/mtd/spi-nor.h
index 4eef4ab0488..b427253b910 100644
--- a/include/linux/mtd/spi-nor.h
+++ b/include/linux/mtd/spi-nor.h
@@ -125,8 +125,6 @@
 /* Used for Macronix and Winbond flashes. */
 #define SPINOR_OP_EN4B		0xb7	/* Enter 4-byte mode */
 #define SPINOR_OP_EX4B		0xe9	/* Exit 4-byte mode */
-#define SPINOR_OP_EN4B			0xb7		/* Enter 4-byte mode */
-#define SPINOR_OP_EX4B			0xe9		/* Exit 4-byte mode */
 #define SPINOR_OP_RD_CR2		0x71		/* Read configuration register 2 */
 #define SPINOR_OP_WR_CR2		0x72		/* Write configuration register 2 */
 #define SPINOR_OP_MXIC_DTR_RD		0xee		/* Fast Read opcode in DTR mode */
-- 
2.43.0

[PATCH v1 2/6] scsi: drop the duplicate COMPARE and SEND DIAGNOSTIC opcodes

[Aristo Chen]

SCSI_COMPARE and SCSI_SEND_DIAG are defined in the table of commands
for all device types and again in the table of commands unique to
direct access devices. SCSI-2 (table 31) lists both as commands for all
device types, so drop the copies from the second table.

The two SCSI_SEND_DIAG copies also disagreed on the command type: the
first said optional, the second mandatory. SCSI-2 names SEND DIAGNOSTIC
as one of the four commands every target implements, together with
INQUIRY, REQUEST SENSE and TEST UNIT READY, so the remaining comment now
says mandatory, as the other three do.

No functional change.

Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 include/scsi.h | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/include/scsi.h b/include/scsi.h
index 2520a8b8fe6..93561749b7f 100644
--- a/include/scsi.h
+++ b/include/scsi.h
@@ -144,13 +144,12 @@ struct scsi_cmd {
 #define SCSI_MODE_SEN10	0x5A		/* Mode Sense 10-byte (Device Specific) */
 #define SCSI_READ_BUFF	0x3C		/* Read Buffer (O) */
 #define SCSI_REQ_SENSE	0x03		/* Request Sense (MANDATORY) */
-#define SCSI_SEND_DIAG	0x1D		/* Send Diagnostic (O) */
+#define SCSI_SEND_DIAG	0x1D		/* Send Diagnostic (MANDATORY) */
 #define SCSI_TST_U_RDY	0x00		/* Test Unit Ready (MANDATORY) */
 #define SCSI_WRITE_BUFF	0x3B		/* Write Buffer (O) */
 /***************************************************************************
  *			  %%% Commands Unique to Direct Access Devices %%%
  ***************************************************************************/
-#define SCSI_COMPARE		0x39		/* Compare (O) */
 #define SCSI_FORMAT		0x04		/* Format Unit (MANDATORY) */
 #define SCSI_LCK_UN_CAC	0x36		/* Lock Unlock Cache (O) */
 #define SCSI_PREFETCH	0x34		/* Prefetch (O) */
@@ -172,7 +171,6 @@ struct scsi_cmd {
 #define SCSI_SRCH_DAT_L	0x32		/* Search Data Low (O) */
 #define SCSI_SEEK6		0x0B		/* Seek 6-Byte (O) */
 #define SCSI_SEEK10		0x2B		/* Seek 10-Byte (O) */
-#define SCSI_SEND_DIAG	0x1D		/* Send Diagnostics (MANDATORY) */
 #define SCSI_SET_LIMIT	0x33		/* Set Limits (O) */
 #define SCSI_START_STP	0x1B		/* Start/Stop Unit (O) */
 #define SCSI_SYNC_CACHE	0x35		/* Synchronize Cache (O) */
-- 
2.43.0

[PATCH v1 3/6] acpi: drop the duplicate ACPI_PPTT_CACHE_TYPE_DATA definition

[Aristo Chen]

ACPI_PPTT_CACHE_TYPE_DATA is defined twice with the same value, at the
start of the cache type values and again after
ACPI_PPTT_CACHE_TYPE_UNIFIED. Drop the second copy.

No functional change.

Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 include/acpi/acpi_table.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/include/acpi/acpi_table.h b/include/acpi/acpi_table.h
index ac50ea10312..d1710d80eee 100644
--- a/include/acpi/acpi_table.h
+++ b/include/acpi/acpi_table.h
@@ -830,7 +830,6 @@ struct acpi_pptt_proc {
 #define ACPI_PPTT_CACHE_TYPE_DATA	0
 #define ACPI_PPTT_CACHE_TYPE_INSTR	1
 #define ACPI_PPTT_CACHE_TYPE_UNIFIED	2
-#define ACPI_PPTT_CACHE_TYPE_DATA	0
 #define ACPI_PPTT_WRITE_THROUGH		BIT(4)
 
 /**
-- 
2.43.0

[PATCH v1 4/6] pci: gt64120: drop the duplicate GT_PCI1M1LD/HD_OFS definitions

[Aristo Chen]

GT_PCI1M1LD_OFS and GT_PCI1M1HD_OFS are defined twice in a row with the
same values. The Linux header this file was imported from,
arch/mips/include/asm/gt64120.h, had the same duplicate, which has now
been dropped in mips-next [1]. Apart from that, Linux has only changed
the header's SPDX tag since the import, so there is nothing else to
re-sync.

Drop the second pair here as well. No functional change.

Link: https://lore.kernel.org/all/20261001053016.76578-1-aristo.chen@canonical.com/ [1]
Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 include/gt64120.h | 2 --
 1 file changed, 2 deletions(-)

diff --git a/include/gt64120.h b/include/gt64120.h
index b8d3f03a5e0..375b702c466 100644
--- a/include/gt64120.h
+++ b/include/gt64120.h
@@ -42,8 +42,6 @@
 #define GT_PCI1M0HD_OFS		0x0a8
 #define GT_PCI1M1LD_OFS		0x0b0
 #define GT_PCI1M1HD_OFS		0x0b8
-#define GT_PCI1M1LD_OFS		0x0b0
-#define GT_PCI1M1HD_OFS		0x0b8
 
 #define GT_SCS10AR_OFS		0x0d0
 #define GT_SCS32AR_OFS		0x0d8
-- 
2.43.0

[PATCH v1 5/6] dt-bindings: soc: imx8_pd: drop the duplicate PD_LVDS0_PWM definition

[Aristo Chen]

PD_LVDS0_PWM is defined twice in a row with the same value. The LVDS1
block lists the same resources with a single PD_LVDS1_PWM, so this is a
plain duplicate, not a typo for a missing LVDS0 resource.

Drop the second copy. No functional change.

Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 include/dt-bindings/soc/imx8_pd.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/include/dt-bindings/soc/imx8_pd.h b/include/dt-bindings/soc/imx8_pd.h
index 682b608eef8..c54056fd9bc 100644
--- a/include/dt-bindings/soc/imx8_pd.h
+++ b/include/dt-bindings/soc/imx8_pd.h
@@ -18,7 +18,6 @@
 #define PD_LVDS0_I2C0               lvds0_i2c0
 #define PD_LVDS0_I2C1               lvds0_i2c1
 #define PD_LVDS0_PWM                lvds0_pwm
-#define PD_LVDS0_PWM                lvds0_pwm
 #define PD_LVDS0_GPIO               lvds0_gpio
 #define PD_DC_1                     dc1_power_domain
 #define PD_DC_1_PLL_0               dc1_pll0
-- 
2.43.0

[PATCH v1 6/6] test: video: define WHITE and SCROLL_LINES once at file scope

[Aristo Chen]

dm_test_video_text() and dm_test_video_text_12x22() each define WHITE
and SCROLL_LINES inside the function body. Macros are not scoped to a
function, so the second pair only redefines the first, and
check_vidconsole_output() further down uses the SCROLL_LINES defined
inside dm_test_video_text().

Define both once at file scope, before their first user, and drop the
two in-function pairs. No functional change.

Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
---
 test/dm/video.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/test/dm/video.c b/test/dm/video.c
index 48ec60c4140..cc854d0e042 100644
--- a/test/dm/video.c
+++ b/test/dm/video.c
@@ -135,15 +135,15 @@ static int video_get_nologo(struct unit_test_state *uts, struct udevice **devp)
 	return 0;
 }
 
+#define WHITE		0xffff
+#define SCROLL_LINES	100
+
 /* Test text output works on the video console */
 static int dm_test_video_text(struct unit_test_state *uts)
 {
 	struct udevice *dev, *con;
 	int i;
 
-#define WHITE		0xffff
-#define SCROLL_LINES	100
-
 	ut_assertok(select_vidconsole(uts, "vidconsole0"));
 	ut_assertok(video_get_nologo(uts, &dev));
 	ut_assertok(uclass_get_device(UCLASS_VIDEO_CONSOLE, 0, &con));
@@ -183,9 +183,6 @@ static int dm_test_video_text_12x22(struct unit_test_state *uts)
 	struct udevice *dev, *con;
 	int i;
 
-#define WHITE		0xffff
-#define SCROLL_LINES	100
-
 	ut_assertok(select_vidconsole(uts, "vidconsole0"));
 	ut_assertok(video_get_nologo(uts, &dev));
 	ut_assertok(uclass_get_device(UCLASS_VIDEO_CONSOLE, 0, &con));
-- 
2.43.0

RE: [PATCH v1 1/6] mtd: spi-nor: drop the duplicate EN4B/EX4B opcode definitions

[Takahiro.Kuwano]

> SPINOR_OP_EN4B and SPINOR_OP_EX4B are defined twice in a row with the
> same values. The second pair came in with the Macronix Octal DTR
> opcodes in commit df3d5f9e410c ("mtd: spi-nor: add support for
> Macronix Octal flash").
> 
> Drop the second pair. No functional change.
> 
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>

Reviewed-by: Takahiro Kuwano <takahiro.kuwano@infineon.com>

Re: [PATCH v1 6/6] test: video: define WHITE and SCROLL_LINES once at file scope

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> test: video: define WHITE and SCROLL_LINES once at file scope
>
> dm_test_video_text() and dm_test_video_text_12x22() each define WHITE
> and SCROLL_LINES inside the function body. Macros are not scoped to a
> function, so the second pair only redefines the first, and
> check_vidconsole_output() further down uses the SCROLL_LINES defined
> inside dm_test_video_text().
>
> Define both once at file scope, before their first user, and drop the
> two in-function pairs. No functional change.
>
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
>
> test/dm/video.c | 9 +++------
>  1 file changed, 3 insertions(+), 6 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>

Re: [PATCH v1 1/6] mtd: spi-nor: drop the duplicate EN4B/EX4B opcode definitions

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> mtd: spi-nor: drop the duplicate EN4B/EX4B opcode definitions
>
> SPINOR_OP_EN4B and SPINOR_OP_EX4B are defined twice in a row with the
> same values. The second pair came in with the Macronix Octal DTR
> opcodes in commit df3d5f9e410c ("mtd: spi-nor: add support for
> Macronix Octal flash").
>
> Drop the second pair. No functional change.
>
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
> Reviewed-by: Takahiro Kuwano <takahiro.kuwano@infineon.com>
>
> include/linux/mtd/spi-nor.h | 2 --
>  1 file changed, 2 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>

Re: [PATCH v1 2/6] scsi: drop the duplicate COMPARE and SEND DIAGNOSTIC opcodes

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> scsi: drop the duplicate COMPARE and SEND DIAGNOSTIC opcodes
>
> SCSI_COMPARE and SCSI_SEND_DIAG are defined in the table of commands
> for all device types and again in the table of commands unique to
> direct access devices. SCSI-2 (table 31) lists both as commands for all
> device types, so drop the copies from the second table.
>
> The two SCSI_SEND_DIAG copies also disagreed on the command type: the
> first said optional, the second mandatory. SCSI-2 names SEND DIAGNOSTIC
> as one of the four commands every target implements, together with
> INQUIRY, REQUEST SENSE and TEST UNIT READY, so the remaining comment now
> says mandatory, as the other three do.
>
> No functional change.
>
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
>
> include/scsi.h | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>

Re: [PATCH v1 5/6] dt-bindings: soc: imx8_pd: drop the duplicate PD_LVDS0_PWM definition

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> dt-bindings: soc: imx8_pd: drop the duplicate PD_LVDS0_PWM definition
>
> PD_LVDS0_PWM is defined twice in a row with the same value. The LVDS1
> block lists the same resources with a single PD_LVDS1_PWM, so this is a
> plain duplicate, not a typo for a missing LVDS0 resource.
>
> Drop the second copy. No functional change.
>
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
>
> include/dt-bindings/soc/imx8_pd.h | 1 -
>  1 file changed, 1 deletion(-)

Reviewed-by: Simon Glass <sjg@chromium.org>

Re: [PATCH v1 4/6] pci: gt64120: drop the duplicate GT_PCI1M1LD/HD_OFS definitions

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> pci: gt64120: drop the duplicate GT_PCI1M1LD/HD_OFS definitions
>
> GT_PCI1M1LD_OFS and GT_PCI1M1HD_OFS are defined twice in a row with the
> same values. The Linux header this file was imported from,
> arch/mips/include/asm/gt64120.h, had the same duplicate, which has now
> been dropped in mips-next [1]. Apart from that, Linux has only changed
> the header's SPDX tag since the import, so there is nothing else to
> re-sync.
>
> Drop the second pair here as well. No functional change.
>
> Link: https://lore.kernel.org/all/20261001053016.76578-1-aristo.chen@canonical.com/ [1]
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
>
> include/gt64120.h | 2 --
>  1 file changed, 2 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>

Re: [PATCH v1 3/6] acpi: drop the duplicate ACPI_PPTT_CACHE_TYPE_DATA definition

[Simon Glass]

On 2026-10-02T07:31:50, Aristo Chen <aristo.chen@canonical.com> wrote:
> acpi: drop the duplicate ACPI_PPTT_CACHE_TYPE_DATA definition
>
> ACPI_PPTT_CACHE_TYPE_DATA is defined twice with the same value, at the
> start of the cache type values and again after
> ACPI_PPTT_CACHE_TYPE_UNIFIED. Drop the second copy.
>
> No functional change.
>
> Signed-off-by: Aristo Chen <aristo.chen@canonical.com>
>
> include/acpi/acpi_table.h | 1 -
>  1 file changed, 1 deletion(-)

Reviewed-by: Simon Glass <sjg@chromium.org>