address all -Wunused-const warnings

From: Arnd Bergmann <[email protected]>

vmpic_msi_feature is only used conditionally, which triggers a rare -Werror=unused-const-variable= warning with gcc:

arch/powerpc/sysdev/fsl_msi.c:567:37: error: 'vmpic_msi_feature' defined but not used [-Werror=unused-const-variable=]
567 | static const struct fsl_msi_feature vmpic_msi_feature =

Hide this one in the same #ifdef as the reference so we can turn on the warning by default.

Fixes 305bcf26128e ("powerpc/fsl-soc: use CONFIG_EPAPR_PARAVIRT for hcalls")

Signed-off-by Arnd Bergmann <[email protected]>

---
  arch/powerpc/sysdev/fsl_msi.c | 2 ++
  1 file changed, 2 insertions(+)

arch/powerpc/sysdev/fsl_msi.c +2 -0

index 8e6c84df4ca1..e205135ae1fe 100644
@@ -564,10 +564,12 @@ static const struct fsl_msi_feature ipic_msi_feature = {
  	.msiir_offset = 0x38,
  };
 
+#ifdef CONFIG_EPAPR_PARAVIRT
  static const struct fsl_msi_feature vmpic_msi_feature = {
  	.fsl_pic_ip = FSL_PIC_IP_VMPIC,
  	.msiir_offset = 0,
  };
+#endif
 
  static const struct of_device_id fsl_of_msi_ids[] = {
  	{

This is one of the drivers with an unused variable that is marked 'const'. Adding a __used annotation here avoids the warning and lets us enable the option by default:

lib/test_ubsan.c:137:28: error: unused variable 'skip_ubsan_array' [-Werror,-Wunused-const-variable]

[...]

From: Arnd Bergmann <[email protected]>

On platforms without ACPI support, this causes a W=1 warning from gcc when the driver is built-in:

drivers/platform/goldfish/goldfish_pipe.c:925:36: warning: 'goldfish_pipe_acpi_match' defined but not used [-Wunused-const-variable=]

  925 | static const struct acpi_device_id goldfish_pipe_acpi_match[] = {
      |                                    ^~~~~~~~~~~~~~~~~~~~~~~~

There is no real harm in just keeping the small table around, so rather than adding more #ifdef checks, just reference it unconditionally to avoid the warning.

From: Arnd Bergmann <[email protected]>

The function using this is hidden in an #ifdef, so the variable needs the same one for a clean W=1 build:

drivers/i2c/busses/i2c-pxa.c:327:26: error: 'icr_bits' defined but not used [-Werror=unused-const-variable=]

Fixes d6a7b5f84b5c ("[ARM] 4827/1: fix two warnings in drivers/i2c/busses/i2c-pxa.c")

Signed-off-by Arnd Bergmann <[email protected]>

From: Krzysztof Kozlowski <[email protected]>

When compile tested with W=1 on x86_64 with driver as built-in:

stmpe-ts.c:371:34: error: unused variable 'stmpe_ts_ids' [-Werror,-Wunused-const-variable]

drivers/input/touchscreen/stmpe-ts.c +0 -0

 	{ .compatible = "st,stmpe-ts", },
 	{ },
 };

Quoted 5 lines · Arnd Bergmann

From: Krzysztof Kozlowski <[email protected]>

When compile tested with W=1 on x86_64 with driver as built-in:

stmpe-ts.c:371:34: error: unused variable 'stmpe_ts_ids' [-Werror,-Wunused-const-variable]

...

drivers/input/touchscreen/stmpe-ts.c static struct platform_driver stmpe_ts_driver = { · line 369

Quoted 2 lines

drivers/input/touchscreen/stmpe-ts.c +1 -1

-static const struct of_device_id stmpe_ts_ids[] = {
+static const struct of_device_id stmpe_ts_ids[] __maybe_unused = {

__maybe_unused?

Why not adding it into .driver as you have done in another patch in this series?

Quoted 1 line · Andy Shevchenko

On Wed, Apr 03, 2024 at 10:06:25AM +0200, Arnd Bergmann wrote:

Quoted 6 lines

drivers/input/touchscreen/stmpe-ts.c static struct platform_driver stmpe_ts_driver = { · line 369

Quoted 2 lines

drivers/input/touchscreen/stmpe-ts.c +1 -1

-static const struct of_device_id stmpe_ts_ids[] = {
+static const struct of_device_id stmpe_ts_ids[] __maybe_unused = {

__maybe_unused?

Why not adding it into .driver as you have done in another patch in this series?

Because there is no benefit in this. This is instantiated by MFD, so the only thing you need is entry for module loading.

Quoted 2 lines · Krzysztof Kozlowski

Quoted 1 line · Andy Shevchenko

On Wed, Apr 03, 2024 at 10:06:25AM +0200, Arnd Bergmann wrote:

...

Quoted 5 lines

Quoted 5 lines

drivers/input/touchscreen/stmpe-ts.c static struct platform_driver stmpe_ts_driver = { · line 369

Quoted 2 lines

drivers/input/touchscreen/stmpe-ts.c +1 -1

-static const struct of_device_id stmpe_ts_ids[] = {
+static const struct of_device_id stmpe_ts_ids[] __maybe_unused = {

__maybe_unused?

Quoted 4 lines

Quoted 1 line

Why not adding it into .driver as you have done in another patch in this series?

Because there is no benefit in this. This is instantiated by MFD, so the only thing you need is entry for module loading.

Hmm... Seems to me rather a good candidate for MODULE_ALIAS in this case. No?

From: Krzysztof Kozlowski <[email protected]>

When compile tested with W=1 on x86_64 with driver as built-in:

stmpe-ts.c:371:34: error: unused variable 'stmpe_ts_ids' [-Werror,-Wunused-const-variable]

Signed-off-by Krzysztof Kozlowski <[email protected]>

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/input/touchscreen/stmpe-ts.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/input/touchscreen/stmpe-ts.c +0 -0

index b204fdb2d22c..022b3e94266d 100644

Quoted 17 lines · Arnd Bergmann

From: Krzysztof Kozlowski <[email protected]>

When compile tested with W=1 on x86_64 with driver as built-in:

stmpe-ts.c:371:34: error: unused variable 'stmpe_ts_ids' [-Werror,-Wunused-const-variable]

Signed-off-by Krzysztof Kozlowski <[email protected]>

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/input/touchscreen/stmpe-ts.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/input/touchscreen/stmpe-ts.c +0 -0

index b204fdb2d22c..022b3e94266d 100644

drivers/input/touchscreen/stmpe-ts.c static struct platform_driver stmpe_ts_driver = { · line 372

Quoted 8 lines

drivers/input/touchscreen/stmpe-ts.c +1 -1

@@ -366,7 +366,7 @@ static struct platform_driver stmpe_ts_driver = {
 };
 module_platform_driver(stmpe_ts_driver);
 
-static const struct of_device_id stmpe_ts_ids[] = {
+static const struct of_device_id stmpe_ts_ids[] __maybe_unused = {
 	{ .compatible = "st,stmpe-ts", },
 	{ },
 };

I'd suggest the following instead:

drivers/input/touchscreen/stmpe-ts.c +10 -9

index b204fdb2d22c..e1afebc641ec 100644
@@ -357,21 +357,22 @@ static void stmpe_ts_remove(struct platform_device *pdev)
 	stmpe_disable(ts->stmpe, STMPE_BLOCK_TOUCHSCREEN);
 }
 
-static struct platform_driver stmpe_ts_driver = {
-	.driver = {
-		.name = STMPE_TS_NAME,
-	},
-	.probe = stmpe_input_probe,
-	.remove_new = stmpe_ts_remove,
-};
-module_platform_driver(stmpe_ts_driver);
-
 static const struct of_device_id stmpe_ts_ids[] = {
 	{ .compatible = "st,stmpe-ts", },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, stmpe_ts_ids);
 
+static struct platform_driver stmpe_ts_driver = {
+	.driver = {
+		.name = STMPE_TS_NAME,
+		.of_match_table = stmpe_ts_ids,
+	},
+	.probe = stmpe_input_probe,
+	.remove_new = stmpe_ts_remove,
+};
+module_platform_driver(stmpe_ts_driver);
+
 MODULE_AUTHOR("Luotao Fu <[email protected]>");
 MODULE_DESCRIPTION("STMPEXXX touchscreen driver");
 MODULE_LICENSE("GPL");

I wonder if with the status quo binding via dt works at all with stmpe_ts_driver.driver.of_match_table unset?!

From: Arnd Bergmann <[email protected]>

With 'make W=1', there is a warning when EFI is turned off but sysfb_efi still gets built:

drivers/firmware/efi/sysfb_efi.c:188:35: error: unused variable 'efifb_dmi_system_table' [-Werror,-Wunused-const-variable]

static const struct dmi_system_id efifb_dmi_system_table[] __initconst = {
                                  ^
drivers/firmware/efi/sysfb_efi.c:238:35: error: unused variable 'efifb_dmi_swap_width_height' [-Werror,-Wunused-const-variable]
static const struct dmi_system_id efifb_dmi_swap_width_height[] __initconst = {
                                  ^

drivers/firmware/efi/sysfb_efi.c:297:28: error: unused function 'find_pci_overlap_node' [-Werror,-Wunused-function] static struct device_node *find_pci_overlap_node(void)

There was an earlier patch to address the duplicate function definitions, but that missed how we should not be building this file in the first place.

Fixes 15d27b15de96 ("efi: sysfb_efi: fix build when EFI is not set")

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/firmware/efi/Makefile    | 3 ++-
 drivers/firmware/efi/sysfb_efi.c | 2 --
 2 files changed, 2 insertions(+), 3 deletions(-)

drivers/firmware/efi/Makefile +2 -1

index a2d0009560d0..3baf80d8cf81 100644
@@ -30,7 +30,8 @@ obj-$(CONFIG_EFI_RCI2_TABLE)          += rci2-table.o
 obj-$(CONFIG_EFI_EMBEDDED_FIRMWARE)    += embedded-firmware.o
 obj-$(CONFIG_LOAD_UEFI_KEYS)           += mokvar-table.o
 
-obj-$(CONFIG_SYSFB)                    += sysfb_efi.o
+sysfb-$(CONFIG_SYSFB)                  += sysfb_efi.o
+obj-$(CONFIG_EFI)                      += $(sysfb-y)
 
 arm-obj-$(CONFIG_EFI)                  := efi-init.o arm-runtime.o
 obj-$(CONFIG_ARM)                      += $(arm-obj-y)

drivers/firmware/efi/sysfb_efi.c +0 -3

index cc807ed35aed..a8c982475f6b 100644
@@ -346,7 +346,6 @@ static const struct fwnode_operations efifb_fwnode_ops = {
        .add_links = efifb_add_links,
 };
 
-#ifdef CONFIG_EFI
 static struct fwnode_handle efifb_fwnode;
 
 __init void sysfb_apply_efi_quirks(void)

@@ -372,4 +371,3 @@ __init void sysfb_set_efifb_fwnode(struct platform_device *pd)
                pd->dev.fwnode = &efifb_fwnode;
        }
 }
-#endif
--

2.39.2

Quoted 15 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

With 'make W=1', there is a warning when EFI is turned off but sysfb_efi still gets built:

drivers/firmware/efi/sysfb_efi.c:188:35: error: unused variable 'efifb_dmi_system_table' [-Werror,-Wunused-const-variable]

static const struct dmi_system_id efifb_dmi_system_table[] __initconst = {
                                  ^
drivers/firmware/efi/sysfb_efi.c:238:35: error: unused variable 'efifb_dmi_swap_width_height' [-Werror,-Wunused-const-variable]
static const struct dmi_system_id efifb_dmi_swap_width_height[] __initconst = {
                                  ^

drivers/firmware/efi/sysfb_efi.c:297:28: error: unused function 'find_pci_overlap_node' [-Werror,-Wunused-function] static struct device_node *find_pci_overlap_node(void)

Quoted 5 lines

There was an earlier patch to address the duplicate function definitions, but that missed how we should not be building this file in the first place.

Fixes 15d27b15de96 ("efi: sysfb_efi: fix build when EFI is not set")

Signed-off-by Arnd Bergmann <[email protected]>

Acked-by Ard Biesheuvel <[email protected]>

or I can take it via the EFI tree if you prefer.

The problem is that the #ifdef checks for some of the structures in this file have gone out of sync with the code referencing them. Update these to match the current usage.

From: Arnd Bergmann <[email protected]>

Building with W=1 shows warnings about unused const variables like this one:

drivers/clk/ti/dpll.c:99:29: error: unused variable 'omap3_dpll_core_ck_ops' [-Werror,-Wunused-const-variable] static const struct clk_ops omap3_dpll_core_ck_ops = {};

The problem is that the #ifdef checks for some of the structures in this file have gone out of sync with the code referencing them. Update these to match the current usage.

Signed-off-by Arnd Bergmann <[email protected]>

---

From: Arnd Bergmann <[email protected]>

The driver_version variable is only used inside of an #ifdef block, which leads to a W=1 warning:

drivers/char/apm-emulation.c:144:19: error: 'driver_version' defined but not used [-Werror=unused-const-variable=]

Move this into the function using it.

Fixes 7726942fb15e ("[APM] Add shared version of APM emulation")

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/char/apm-emulation.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

drivers/char/apm-emulation.c +0 -3

index e795390b070f..53ce352f7197 100644
@@ -141,9 +141,6 @@ static struct apm_queue kapmd_queue;
 
 static DEFINE_MUTEX(state_lock);
 
-static const char driver_version[] = "1.13";	/* no spaces */
-
-
 
 /*
  * Compatibility cruft until the IPAQ people move over to the new

From: Arnd Bergmann <[email protected]>

Building with W=1 shows that a couple of variables in this driver are only used in certain configurations:

drivers/video/fbdev/sis/init301.c:239:28: error: 'SiS_Part2CLVX_6' defined but not used [-Werror=unused-const-variable=]
...

Fixes 5908986ef348 ("video: fbdev: sis: avoid mismatched prototypes")

Signed-off-by Arnd Bergmann <[email protected]>

---
  drivers/video/fbdev/sis/init301.c | 3 +--
  1 file changed, 1 insertion(+), 2 deletions(-)

Building with W=1 shows a warning about an unused dmi_system_id table:

drivers/leds/leds-apu.c:85:35: error: 'apu_led_dmi_table' defined but not used [-Werror=unused-const-variable=]
85 | static const struct dmi_system_id apu_led_dmi_table[] __initconst = {

Since the current version doesn't even do anything about the different implementations but only checks the type of system, just drop the custom lookup logic and call dmi_check_system() using the table itself.

[...]

From: Arnd Bergmann <[email protected]>

When the driver is built-in, 'make W=1' warns about an unused ID table:

drivers/iio/dac/ad5755.c:866:34: error: 'ad5755_of_match' defined but not used [-Werror=unused-const-variable=]
866 | static const struct of_device_id ad5755_of_match[] = {

While the data is duplicated in the spi_device_id, it's common to use the actual OF compatible strings in the driver.

Since there are no in-tree users of plain platform devices, the spi_device_id table could actually be dropped entirely with this.

From: Arnd Bergmann <[email protected]>

When the driver is built-in, 'make W=1' warns about an unused ID table:

drivers/iio/dac/ad5755.c:866:34: error: 'ad5755_of_match' defined but not used [-Werror=unused-const-variable=]
866 | static const struct of_device_id ad5755_of_match[] = {

While the data is duplicated in the spi_device_id, it's common to use the actual OF compatible strings in the driver.

Since there are no in-tree users of plain platform devices, the spi_device_id table could actually be dropped entirely with this.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/iio/dac/ad5755.c | 1 +
 1 file changed, 1 insertion(+)

drivers/iio/dac/ad5755.c +0 -0

index 404865e35460..5c1e7f428c25 100644

Quoted 24 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

When the driver is built-in, 'make W=1' warns about an unused ID table:

drivers/iio/dac/ad5755.c:866:34: error: 'ad5755_of_match' defined but not used [-Werror=unused-const-variable=]
866 | static const struct of_device_id ad5755_of_match[] = {

While the data is duplicated in the spi_device_id, it's common to use the actual OF compatible strings in the driver.

Since there are no in-tree users of plain platform devices, the spi_device_id table could actually be dropped entirely with this.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/iio/dac/ad5755.c | 1 +
 1 file changed, 1 insertion(+)

drivers/iio/dac/ad5755.c +0 -0

index 404865e35460..5c1e7f428c25 100644

drivers/iio/dac/ad5755.c MODULE_DEVICE_TABLE(of, ad5755_of_match); · line 879

Quoted 4 lines

drivers/iio/dac/ad5755.c +1 -0

@@ -876,6 +876,7 @@ MODULE_DEVICE_TABLE(of, ad5755_of_match);
 static struct spi_driver ad5755_driver = {
 	.driver = {
 		.name = "ad5755",
+		.of_match_table = ad5755_of_match,

I was working on this as well and have a bit bigger solution, following Jonathan's preference (I think):

https://lore.kernel.org/all/20240226192555.14aa178e@jic23-huawei/

I need to send v3, somehow I missed his comments.

Quoted 33 lines · Krzysztof Kozlowski

Quoted 24 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

When the driver is built-in, 'make W=1' warns about an unused ID table:

drivers/iio/dac/ad5755.c:866:34: error: 'ad5755_of_match' defined but not used [-Werror=unused-const-variable=]
866 | static const struct of_device_id ad5755_of_match[] = {

While the data is duplicated in the spi_device_id, it's common to use the actual OF compatible strings in the driver.

Since there are no in-tree users of plain platform devices, the spi_device_id table could actually be dropped entirely with this.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/iio/dac/ad5755.c | 1 +
 1 file changed, 1 insertion(+)

drivers/iio/dac/ad5755.c +0 -0

index 404865e35460..5c1e7f428c25 100644

drivers/iio/dac/ad5755.c MODULE_DEVICE_TABLE(of, ad5755_of_match); · line 879

Quoted 4 lines

drivers/iio/dac/ad5755.c +1 -0

@@ -876,6 +876,7 @@ MODULE_DEVICE_TABLE(of, ad5755_of_match);
 static struct spi_driver ad5755_driver = {
 	.driver = {
 		.name = "ad5755",
+		.of_match_table = ad5755_of_match,  

I was working on this as well and have a bit bigger solution, following Jonathan's preference (I think):

Quoted 3 lines

https://lore.kernel.org/all/20240226192555.14aa178e@jic23-huawei/

I need to send v3, somehow I missed his comments.

Yes, that looks good as well, though you might need to drop spi_device_id table if you convert it to using pointers.

Arnd

Quoted 24 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

When the driver is built-in, 'make W=1' warns about an unused ID table:

drivers/iio/dac/ad5755.c:866:34: error: 'ad5755_of_match' defined but not used [-Werror=unused-const-variable=]
866 | static const struct of_device_id ad5755_of_match[] = {

While the data is duplicated in the spi_device_id, it's common to use the actual OF compatible strings in the driver.

Since there are no in-tree users of plain platform devices, the spi_device_id table could actually be dropped entirely with this.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/iio/dac/ad5755.c | 1 +
 1 file changed, 1 insertion(+)

drivers/iio/dac/ad5755.c +0 -0

index 404865e35460..5c1e7f428c25 100644

drivers/iio/dac/ad5755.c MODULE_DEVICE_TABLE(of, ad5755_of_match); · line 879

Quoted 4 lines

drivers/iio/dac/ad5755.c +1 -0

@@ -876,6 +876,7 @@ MODULE_DEVICE_TABLE(of, ad5755_of_match);
 static struct spi_driver ad5755_driver = {
 	.driver = {
 		.name = "ad5755",
+		.of_match_table = ad5755_of_match,  

I was working on this as well and have a bit bigger solution, following Jonathan's preference (I think):

https://lore.kernel.org/all/20240226192555.14aa178e@jic23-huawei/

I need to send v3, somehow I missed his comments.

Jonathan,
Do you want me to still work on this according to your comments (which I missed, I am sorry).

Best regards,
Krzysztof

From: Arnd Bergmann <[email protected]>

The arche-ctrl has two platform drivers and three of_device_id tables, but one table is only used for the the module loader, while the other two seem to be associated with their drivers.

This leads to a W=1 warning when the driver is built-in:

drivers/staging/greybus/arche-platform.c:623:34: error: 'arche_combined_id' defined but not used [-Werror=unused-const-variable=]
623 | static const struct of_device_id arche_combined_id[] = {

Drop the extra table and register both tables that are actually used as the ones for the module loader instead.

Fixes 7b62b61c752a ("greybus: arche-ctrl: Don't expose driver internals to arche-platform driver")

Signed-off-by Arnd Bergmann <[email protected]>

---
  drivers/staging/greybus/arche-apb-ctrl.c | 1 +
  drivers/staging/greybus/arche-platform.c | 9 +--------
  2 files changed, 2 insertions(+), 8 deletions(-)

drivers/staging/greybus/arche-apb-ctrl.c +1 -0

index 8541995008da..aa6f266b62a1 100644
@@ -466,6 +466,7 @@ static const struct of_device_id arche_apb_ctrl_of_match[] = {
  	{ .compatible = "usbffff,2", },
  	{ },
  };
+MODULE_DEVICE_TABLE(of, arche_apb_ctrl_of_match);
 
  static struct platform_driver arche_apb_ctrl_device_driver = {
  	.probe		= arche_apb_ctrl_probe,

drivers/staging/greybus/arche-platform.c +1 -8

index 891b75327d7f..b33977ccd527 100644
@@ -619,14 +619,7 @@ static const struct of_device_id arche_platform_of_match[] = {
  	{ .compatible = "google,arche-platform", },
  	{ },
  };
-
-static const struct of_device_id arche_combined_id[] = {
-	/* Use PID/VID of SVC device */
-	{ .compatible = "google,arche-platform", },
-	{ .compatible = "usbffff,2", },
-	{ },
-};
-MODULE_DEVICE_TABLE(of, arche_combined_id);
+MODULE_DEVICE_TABLE(of, arche_platform_of_match);
 
  static struct platform_driver arche_platform_device_driver = {
  	.probe		= arche_platform_probe,

From: Arnd Bergmann <[email protected]>

When CONFIG_NET is disabled, an extra warning shows up for this unused variable:

lib/checksum_kunit.c:218:18: error: 'expected_csum_ipv6_magic' defined but not used [-Werror=unused-const-variable=]

Hide it under the same #ifdef as the reference to it.

Fixes f24a70106dc1 ("lib: checksum: Fix build with CONFIG_NET=n")

Signed-off-by Arnd Bergmann <[email protected]>

---
  lib/checksum_kunit.c | 2 ++
  1 file changed, 2 insertions(+)

lib/checksum_kunit.c +2 -0

index bf70850035c7..80dd1e1b71ba 100644
@@ -215,6 +215,7 @@ static const u32 init_sums_no_overflow[] = {
  	0xffff0000, 0xfffffffb,
  };
 
+#ifdef CONFIG_NET
  static const u16 expected_csum_ipv6_magic[] = {
  	0x18d4, 0x3085, 0x2e4b, 0xd9f4, 0xbdc8, 0x78f,	0x1034, 0x8422, 0x6fc0,
  	0xd2f6, 0xbeb5, 0x9d3,	0x7e2a, 0x312e, 0x778e, 0xc1bb, 0x7cf2, 0x9d1e,

@@ -240,6 +241,7 @@ static const u16 expected_csum_ipv6_magic[] = {
  	0x99aa, 0xb06b, 0xee19, 0xcc2c, 0xf34c, 0x7c49, 0xdac3, 0xa71e, 0xc988,
  	0x3845, 0x1014
  };
+#endif
 
  static const u16 expected_fast_csum[] = {
  	0xda83, 0x45da, 0x4f46, 0x4e4f, 0x34e,	0xe902, 0xa5e9, 0x87a5, 0x7187,

Le 03/04/2024 à 10:06, Arnd Bergmann a écrit :

Quoted 10 lines

From: Arnd Bergmann <[email protected]>

When CONFIG_NET is disabled, an extra warning shows up for this unused variable:

lib/checksum_kunit.c:218:18: error: 'expected_csum_ipv6_magic' defined but not used [-Werror=unused-const-variable=]

Hide it under the same #ifdef as the reference to it.

Fixes f24a70106dc1 ("lib: checksum: Fix build with CONFIG_NET=n")

I think that commit introduced unjustified #ifdef in a C file.

Refer
https://docs.kernel.org/process/coding-style.html#conditional-compilation

The fix should be:

(quoted diff) +3 -0

+	if (!IS_ENABLED(CONFIG_NET))
+		return;
+
  	for (int i = 0; i < NUM_IPv6_TESTS; i++) {
  		saddr = (const struct in6_addr *)(random_buf + i);
  		daddr = (const struct in6_addr *)(random_buf + i +

Quoted 19 lines

Le 03/04/2024 à 10:06, Arnd Bergmann a écrit :

Quoted 10 lines

From: Arnd Bergmann <[email protected]>

When CONFIG_NET is disabled, an extra warning shows up for this unused variable:

lib/checksum_kunit.c:218:18: error: 'expected_csum_ipv6_magic' defined but not used [-Werror=unused-const-variable=]

Hide it under the same #ifdef as the reference to it.

Fixes f24a70106dc1 ("lib: checksum: Fix build with CONFIG_NET=n")

I think that commit introduced unjustified #ifdef in a C file.

Refer
https://docs.kernel.org/process/coding-style.html#conditional-compilation

The fix should be:

Review static void test_csum_ipv6_magic(struct kunit *test) · line 621

Quoted 11 lines

(quoted diff) +3 -1

+	if (!IS_ENABLED(CONFIG_NET))
+		return;
+
   	for (int i = 0; i < NUM_IPv6_TESTS; i++) {
   		saddr = (const struct in6_addr *)(random_buf + i);
   		daddr = (const struct in6_addr *)(random_buf + i +

@@ -618,7 +620,6 @@ static void test_csum_ipv6_magic(struct kunit *test)
   		CHECK_EQ(to_sum16(expected_csum_ipv6_magic[i]),
   			 csum_ipv6_magic(saddr, daddr, len, proto, csum));
   	}
-#endif /* !CONFIG_NET */

Yes, I usually prefer that approach, and I think I tried it first but ran into compile-time problems because csum_ipv6_magic() is declared conditionally. Let me try again, maybe I misremember what I did or I can remove some more #ifdef checks for this.

Le 03/04/2024 à 10:58, Arnd Bergmann a écrit :

Quoted 5 lines · Christophe Leroy

Yes, I usually prefer that approach, and I think I tried it first but ran into compile-time problems because csum_ipv6_magic() is declared conditionally. Let me try again, maybe I misremember what I did or I can remove some more #ifdef checks for this.

FWIW, before replying I did a build check with pmac32_defconfig (powerpc) with CONFIG_NET removed. I was able to build lib/checksum_kunit.o and objdump shown there was no reference to csum_ipv6_magic() inside it.

From: Arnd Bergmann <[email protected]>

There is a warning about unused variables when building with W=1 and no procfs:

net/sunrpc/cache.c:1660:30: error: 'cache_flush_proc_ops' defined but not used [-Werror=unused-const-variable=]

 1660 | static const struct proc_ops cache_flush_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~~~~~
net/sunrpc/cache.c:1622:30: error: 'content_proc_ops' defined but not used [-Werror=unused-const-variable=]
 1622 | static const struct proc_ops content_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~
net/sunrpc/cache.c:1598:30: error: 'cache_channel_proc_ops' defined but not used [-Werror=unused-const-variable=]
 1598 | static const struct proc_ops cache_channel_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~~~~~~~

These are used inside of an #ifdef, so replacing that with an IS_ENABLED() check lets the compiler see how they are used while still dropping them during dead code elimination.

Fixes dbf847ecb631 ("knfsd: allow cache_register to return error on failure")

Signed-off-by Arnd Bergmann <[email protected]>

---
 net/sunrpc/cache.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

net/sunrpc/cache.c +3 -1

index 95ff74706104..ab3a57965dc0 100644
@@ -1673,12 +1673,14 @@ static void remove_cache_proc_entries(struct cache_detail *cd)
 	}
 }
 
-#ifdef CONFIG_PROC_FS
 static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
 {
 	struct proc_dir_entry *p;
 	struct sunrpc_net *sn;
 
+	if (!IS_ENABLED(CONFIG_PROC_FS))
+		return 0;
+
 	sn = net_generic(net, sunrpc_net_id);
 	cd->procfs = proc_mkdir(cd->name, sn->proc_net_rpc);
 	if (cd->procfs == NULL)

@@ -1706,12 +1708,6 @@ static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
 	remove_cache_proc_entries(cd);
 	return -ENOMEM;

From: Arnd Bergmann <[email protected]>

There is a warning about unused variables when building with W=1 and no procfs:

net/sunrpc/cache.c:1660:30: error: 'cache_flush_proc_ops' defined but not used [-Werror=unused-const-variable=]

 1660 | static const struct proc_ops cache_flush_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~~~~~
net/sunrpc/cache.c:1622:30: error: 'content_proc_ops' defined but not used [-Werror=unused-const-variable=]
 1622 | static const struct proc_ops content_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~
net/sunrpc/cache.c:1598:30: error: 'cache_channel_proc_ops' defined but not used [-Werror=unused-const-variable=]
 1598 | static const struct proc_ops cache_channel_proc_ops = {
      |                              ^~~~~~~~~~~~~~~~~~~~~~

These are used inside of an #ifdef, so replacing that with an IS_ENABLED() check lets the compiler see how they are used while still dropping them during dead code elimination.

Fixes dbf847ecb631 ("knfsd: allow cache_register to return error on failure")

Signed-off-by Arnd Bergmann <[email protected]>

---
 net/sunrpc/cache.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

net/sunrpc/cache.c +3 -7

index 95ff74706104..ab3a57965dc0 100644
@@ -1673,12 +1673,14 @@ static void remove_cache_proc_entries(struct cache_detail *cd)
 	}
 }
 
-#ifdef CONFIG_PROC_FS
 static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
 {
 	struct proc_dir_entry *p;
 	struct sunrpc_net *sn;
 
+	if (!IS_ENABLED(CONFIG_PROC_FS))
+		return 0;
+
 	sn = net_generic(net, sunrpc_net_id);
 	cd->procfs = proc_mkdir(cd->name, sn->proc_net_rpc);
 	if (cd->procfs == NULL)

@@ -1706,12 +1708,6 @@ static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
 	remove_cache_proc_entries(cd);
 	return -ENOMEM;
 }
-#else /* CONFIG_PROC_FS */
-static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
-{
-	return 0;
-}
-#endif
 
 void __init cache_initialize(void)
 {

Signature · 1 line

2.39.2

From: Arnd Bergmann <[email protected]>

The reference to il_rate_mcs is inside of an #ifdef, causing a W=1 warning:

drivers/net/wireless/intel/iwlegacy/4965-rs.c:189:38: error: unused variable 'il_rate_mcs' [-Werror,-Wunused-const-variable] static const struct il_rate_mcs_info il_rate_mcs[RATE_COUNT] = {

Replace the #ifdef with a PTR_IF() for better compile time analysis. The dead code will still get eliminated, but the warning goes away.

From: Arnd Bergmann <[email protected]>

The reference to il_rate_mcs is inside of an #ifdef, causing a W=1 warning:

drivers/net/wireless/intel/iwlegacy/4965-rs.c:189:38: error: unused variable 'il_rate_mcs' [-Werror,-Wunused-const-variable] static const struct il_rate_mcs_info il_rate_mcs[RATE_COUNT] = {

Replace the #ifdef with a PTR_IF() for better compile time analysis. The dead code will still get eliminated, but the warning goes away.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/net/wireless/intel/iwlegacy/4965-rs.c | 15 ++-------------
 drivers/net/wireless/intel/iwlegacy/common.h  |  2 --
 2 files changed, 2 insertions(+), 15 deletions(-)

drivers/net/wireless/intel/iwlegacy/4965-rs.c +0 -7

index 718efb1aa1b0..1aa2cee5d131 100644
@@ -132,15 +132,8 @@ static void il4965_rs_fill_link_cmd(struct il_priv *il,
 static void il4965_rs_stay_in_table(struct il_lq_sta *lq_sta,
 				    bool force_search);
 
-#ifdef CONFIG_MAC80211_DEBUGFS
 static void il4965_rs_dbgfs_set_mcs(struct il_lq_sta *lq_sta,
 				    u32 *rate_n_flags, int idx);
-#else
-static void
-il4965_rs_dbgfs_set_mcs(struct il_lq_sta *lq_sta, u32 * rate_n_flags, int idx)
-{
-}
-#endif
 
 /*
  * The following tables contain the expected throughput metrics for all rates

drivers/net/wireless/intel/iwlegacy/4965-rs.c +0 -1

 static void
 il4965_rs_dbgfs_set_mcs(struct il_lq_sta *lq_sta, u32 * rate_n_flags, int idx)
 {

@@ -2758,7 +2749,6 @@ il4965_rs_add_debugfs(void *il, void *il_sta, struct dentry *dir)
 	debugfs_create_u8("tx_agg_tid_enable", 0600, dir,
 			  &lq_sta->tx_agg_tid_en);
 }
-#endif
 
 /*
  * Initialization of rate scaling information is done by driver after

drivers/net/wireless/intel/iwlegacy/4965-rs.c +0 -0

 };
 
 int

drivers/net/wireless/intel/iwlegacy/common.h +0 -0

index 69687fcf963f..b9f1daf0901b 100644

drivers/net/wireless/intel/iwlegacy/common.h +0 -0

 	struct il_priv *drv;
 
 	/* used to be in sta_info */

From: Arnd Bergmann <[email protected]>

When CONFIG_OF is disabled, there is no reference to this variable:

drivers/firmware/qcom/qcom_scm.c:1655:34: error: unused variable 'qcom_scm_qseecom_allowlist' [-Werror,-Wunused-const-variable] static const struct of_device_id qcom_scm_qseecom_allowlist[] = {

Mark it as __maybe_unused to hide this warning.

Fixes 00b1248606ba ("firmware: qcom_scm: Add support for Qualcomm Secure Execution Environment SCM interface")

Signed-off-by Arnd Bergmann <[email protected]>

---

From: Arnd Bergmann <[email protected]>

When both ACPI and OF are disabled, the dev_vdata variable is unused:

drivers/crypto/ccp/sp-platform.c:33:34: error: unused variable 'dev_vdata' [-Werror,-Wunused-const-variable]

This is not a useful configuration, and there is not much point in saving a few bytes when only one of the two is enabled, so just remove all these ifdef checks and rely on of_match_node() and acpi_match_device() returning NULL when these subsystems are disabled.

Fixes 6c5063434098 ("crypto: ccp - Add ACPI support")

Signed-off-by Arnd Bergmann <[email protected]>

---
  drivers/crypto/ccp/sp-platform.c | 14 ++------------
  1 file changed, 2 insertions(+), 12 deletions(-)

drivers/crypto/ccp/sp-platform.c +2 -12

index 473301237760..ff6ceb4feee0 100644
@@ -39,44 +39,38 @@ static const struct sp_dev_vdata dev_vdata[] = {
  	},
  };
 
-#ifdef CONFIG_ACPI
  static const struct acpi_device_id sp_acpi_match[] = {
  	{ "AMDI0C00", (kernel_ulong_t)&dev_vdata[0] },
  	{ },
  };
  MODULE_DEVICE_TABLE(acpi, sp_acpi_match);
-#endif
 
-#ifdef CONFIG_OF
  static const struct of_device_id sp_of_match[] = {
  	{ .compatible = "amd,ccp-seattle-v1a",
  	  .data = (const void *)&dev_vdata[0] },
  	{ },
  };
  MODULE_DEVICE_TABLE(of, sp_of_match);
-#endif
 
  static struct sp_dev_vdata *sp_get_of_version(struct platform_device *pdev)
  {
-#ifdef CONFIG_OF
  	const struct of_device_id *match;
 
  	match = of_match_node(sp_of_match, pdev->dev.of_node);
  	if (match && match->data)
  		return (struct sp_dev_vdata *)match->data;
-#endif
+
  	return NULL;
  }
 
  static struct sp_dev_vdata *sp_get_acpi_version(struct platform_device *pdev)
  {
-#ifdef CONFIG_ACPI
  	const struct acpi_device_id *match;
 
  	match = acpi_match_device(sp_acpi_match, &pdev->dev);
  	if (match && match->driver_data)
  		return (struct sp_dev_vdata *)match->driver_data;
-#endif
+
  	return NULL;
  }
 

@@ -212,12 +206,8 @@ static int sp_platform_resume(struct platform_device *pdev)
  static struct platform_driver sp_platform_driver = {
  	.driver = {
  		.name = "ccp",
-#ifdef CONFIG_ACPI
  		.acpi_match_table = sp_acpi_match,
-#endif
-#ifdef CONFIG_OF
  		.of_match_table = sp_of_match,
-#endif
  	},
  	.probe = sp_platform_probe,
  	.remove_new = sp_platform_remove,

Quoted 11 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

When both ACPI and OF are disabled, the dev_vdata variable is unused:

drivers/crypto/ccp/sp-platform.c:33:34: error: unused variable 'dev_vdata' [-Werror,-Wunused-const-variable]

This is not a useful configuration, and there is not much point in saving a few bytes when only one of the two is enabled, so just remove all these ifdef checks and rely on of_match_node() and acpi_match_device() returning NULL when these subsystems are disabled.

Quoted 2 lines

Fixes 6c5063434098 ("crypto: ccp - Add ACPI support")

Signed-off-by Arnd Bergmann <[email protected]>

Would using __maybe_unused on dev_vdata be the safer, easier choice?

From: Arnd Bergmann <[email protected]>

When both ACPI and OF are disabled, the dev_vdata variable is unused:

drivers/crypto/ccp/sp-platform.c:33:34: error: unused variable 'dev_vdata' [-Werror,-Wunused-const-variable]

This is not a useful configuration, and there is not much point in saving a few bytes when only one of the two is enabled, so just remove all these ifdef checks and rely on of_match_node() and acpi_match_device() returning NULL when these subsystems are disabled.

Fixes 6c5063434098 ("crypto: ccp - Add ACPI support")

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/crypto/ccp/sp-platform.c | 14 ++------------
 1 file changed, 2 insertions(+), 12 deletions(-)

From: Arnd Bergmann <[email protected]>

The driver_desc variable is only used in some configurations:

drivers/usb/gadget/udc/omap_udc.c:113:19: error: unused variable 'driver_desc' [-Werror,-Wunused-const-variable]

Since there is only ever one user of it, just open-code the string in place and remove the global variable and the macro behind it. This also helps grep for the MODULE_DESCRIPTION string.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/usb/gadget/udc/omap_udc.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

drivers/usb/gadget/udc/omap_udc.c +3 -7

index f90eeecf27de..e13b8ec8ef8a 100644
@@ -56,7 +56,6 @@
 /* ISO too */
 #define	USE_ISO
 
-#define	DRIVER_DESC	"OMAP UDC driver"
 #define	DRIVER_VERSION	"4 October 2004"
 
 #define OMAP_DMA_USB_W2FC_TX0		29

@@ -110,7 +109,6 @@ MODULE_PARM_DESC(use_dma, "enable/disable DMA");
 
 
 static const char driver_name[] = "omap_udc";
-static const char driver_desc[] = DRIVER_DESC;
 
 /*-------------------------------------------------------------------------*/
 

@@ -2299,13 +2297,11 @@ static int proc_udc_show(struct seq_file *s, void *_)
 
 	spin_lock_irqsave(&udc->lock, flags);
 
-	seq_printf(s, "%s, version: " DRIVER_VERSION
+	seq_printf(s, "OMAP UDC driver, version: " DRIVER_VERSION
 #ifdef	USE_ISO
 		" (iso)"
 #endif
-		"%s\n",
-		driver_desc,
-		use_dma ?  " (dma)" : "");
+		"%s\n", use_dma ?  " (dma)" : "");
 
 	tmp = omap_readw(UDC_REV) & 0xff;
 	seq_printf(s,

@@ -2994,6 +2990,6 @@ static struct platform_driver udc_driver = {
 
 module_platform_driver(udc_driver);
 
-MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_DESCRIPTION("OMAP UDC driver");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("platform:omap_udc");

Signature · 1 line

2.39.2

From: Arnd Bergmann <[email protected]>

The reference to this variable is hidden in an #ifdef:

drivers/cpufreq/intel_pstate.c:2440:32: error: 'intel_pstate_cpu_oob_ids' defined but not used [-Werror=unused-const-variable=]

Use the same check around the definition.

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/cpufreq/intel_pstate.c | 2 ++
 1 file changed, 2 insertions(+)

drivers/cpufreq/intel_pstate.c +1 -0

index dbbf299f4219..29ce9edc6f68 100644
@@ -2437,6 +2437,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
 };
 MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids);
 
+#ifdef CONFIG_ACPI
 static const struct x86_cpu_id intel_pstate_cpu_oob_ids[] __initconst = {
        X86_MATCH(BROADWELL_D,          core_funcs),
        X86_MATCH(BROADWELL_X,          core_funcs),

From: Arnd Bergmann <[email protected]>

Building with W=1 shows a warning for an unused variable when CONFIG_PCI is diabled:

drivers/ata/sata_mv.c:790:35: error: unused variable 'mv_pci_tbl' [-Werror,-Wunused-const-variable] static const struct pci_device_id mv_pci_tbl[] = {

Move the table into the same block that containsn the pci_driver definition.

Fixes 7bb3c5290ca0 ("sata_mv: Remove PCI dependency")

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/ata/sata_mv.c | 64 +++++++++++++++++++++----------------------
 1 file changed, 32 insertions(+), 32 deletions(-)

drivers/ata/sata_mv.c +32 -32

index e82786c63fbd..697063890f5d 100644
@@ -787,37 +787,6 @@ static const struct ata_port_info mv_port_info[] = {
 	},
 };
 
-static const struct pci_device_id mv_pci_tbl[] = {
-	{ PCI_VDEVICE(MARVELL, 0x5040), chip_504x },
-	{ PCI_VDEVICE(MARVELL, 0x5041), chip_504x },
-	{ PCI_VDEVICE(MARVELL, 0x5080), chip_5080 },
-	{ PCI_VDEVICE(MARVELL, 0x5081), chip_508x },
-	/* RocketRAID 1720/174x have different identifiers */
-	{ PCI_VDEVICE(TTI, 0x1720), chip_6042 },
-	{ PCI_VDEVICE(TTI, 0x1740), chip_6042 },
-	{ PCI_VDEVICE(TTI, 0x1742), chip_6042 },
-
-	{ PCI_VDEVICE(MARVELL, 0x6040), chip_604x },
-	{ PCI_VDEVICE(MARVELL, 0x6041), chip_604x },
-	{ PCI_VDEVICE(MARVELL, 0x6042), chip_6042 },
-	{ PCI_VDEVICE(MARVELL, 0x6080), chip_608x },
-	{ PCI_VDEVICE(MARVELL, 0x6081), chip_608x },
-
-	{ PCI_VDEVICE(ADAPTEC2, 0x0241), chip_604x },
-
-	/* Adaptec 1430SA */
-	{ PCI_VDEVICE(ADAPTEC2, 0x0243), chip_7042 },
-
-	/* Marvell 7042 support */
-	{ PCI_VDEVICE(MARVELL, 0x7042), chip_7042 },
-
-	/* Highpoint RocketRAID PCIe series */
-	{ PCI_VDEVICE(TTI, 0x2300), chip_7042 },
-	{ PCI_VDEVICE(TTI, 0x2310), chip_7042 },
-
-	{ }			/* terminate list */
-};
-
 static const struct mv_hw_ops mv5xxx_ops = {
 	.phy_errata		= mv5_phy_errata,
 	.enable_leds		= mv5_enable_leds,

@@ -4303,6 +4272,37 @@ static int mv_pci_init_one(struct pci_dev *pdev,
 static int mv_pci_device_resume(struct pci_dev *pdev);
 #endif
 
+static const struct pci_device_id mv_pci_tbl[] = {
+	{ PCI_VDEVICE(MARVELL, 0x5040), chip_504x },
+	{ PCI_VDEVICE(MARVELL, 0x5041), chip_504x },
+	{ PCI_VDEVICE(MARVELL, 0x5080), chip_5080 },
+	{ PCI_VDEVICE(MARVELL, 0x5081), chip_508x },
+	/* RocketRAID 1720/174x have different identifiers */
+	{ PCI_VDEVICE(TTI, 0x1720), chip_6042 },
+	{ PCI_VDEVICE(TTI, 0x1740), chip_6042 },
+	{ PCI_VDEVICE(TTI, 0x1742), chip_6042 },
+
+	{ PCI_VDEVICE(MARVELL, 0x6040), chip_604x },
+	{ PCI_VDEVICE(MARVELL, 0x6041), chip_604x },
+	{ PCI_VDEVICE(MARVELL, 0x6042), chip_6042 },
+	{ PCI_VDEVICE(MARVELL, 0x6080), chip_608x },
+	{ PCI_VDEVICE(MARVELL, 0x6081), chip_608x },
+
+	{ PCI_VDEVICE(ADAPTEC2, 0x0241), chip_604x },
+
+	/* Adaptec 1430SA */
+	{ PCI_VDEVICE(ADAPTEC2, 0x0243), chip_7042 },
+
+	/* Marvell 7042 support */
+	{ PCI_VDEVICE(MARVELL, 0x7042), chip_7042 },
+
+	/* Highpoint RocketRAID PCIe series */
+	{ PCI_VDEVICE(TTI, 0x2300), chip_7042 },
+	{ PCI_VDEVICE(TTI, 0x2310), chip_7042 },
+
+	{ }			/* terminate list */
+};
+
 
 static struct pci_driver mv_pci_driver = {
 	.name			= DRV_NAME,

@@ -4315,6 +4315,7 @@ static struct pci_driver mv_pci_driver = {
 #endif
 
 };
+MODULE_DEVICE_TABLE(pci, mv_pci_tbl);
 
 /**
  *      mv_print_info - Dump key info to kernel log for perusal.

@@ -4487,7 +4488,6 @@ static void __exit mv_exit(void)
 MODULE_AUTHOR("Brett Russ");
 MODULE_DESCRIPTION("SCSI low-level driver for Marvell SATA controllers");
 MODULE_LICENSE("GPL v2");
-MODULE_DEVICE_TABLE(pci, mv_pci_tbl);
 MODULE_VERSION(DRV_VERSION);
 MODULE_ALIAS("platform:" DRV_NAME);
 

From: Arnd Bergmann <[email protected]>

Building with W=1 shows a warning for an unused variable when CONFIG_PCI is diabled:

drivers/ata/sata_mv.c:790:35: error: unused variable 'mv_pci_tbl' [-Werror,-Wunused-const-variable] static const struct pci_device_id mv_pci_tbl[] = {

Move the table into the same block that containsn the pci_driver definition.

Fixes 7bb3c5290ca0 ("sata_mv: Remove PCI dependency")

Signed-off-by Arnd Bergmann <[email protected]>

---
 drivers/ata/sata_mv.c | 64 +++++++++++++++++++++----------------------
 1 file changed, 32 insertions(+), 32 deletions(-)

drivers/ata/sata_mv.c +32 -32

index e82786c63fbd..697063890f5d 100644
@@ -787,37 +787,6 @@ static const struct ata_port_info mv_port_info[] = {
 	},
 };
 
-static const struct pci_device_id mv_pci_tbl[] = {
-	{ PCI_VDEVICE(MARVELL, 0x5040), chip_504x },
-	{ PCI_VDEVICE(MARVELL, 0x5041), chip_504x },
-	{ PCI_VDEVICE(MARVELL, 0x5080), chip_5080 },
-	{ PCI_VDEVICE(MARVELL, 0x5081), chip_508x },
-	/* RocketRAID 1720/174x have different identifiers */
-	{ PCI_VDEVICE(TTI, 0x1720), chip_6042 },
-	{ PCI_VDEVICE(TTI, 0x1740), chip_6042 },
-	{ PCI_VDEVICE(TTI, 0x1742), chip_6042 },
-
-	{ PCI_VDEVICE(MARVELL, 0x6040), chip_604x },
-	{ PCI_VDEVICE(MARVELL, 0x6041), chip_604x },
-	{ PCI_VDEVICE(MARVELL, 0x6042), chip_6042 },
-	{ PCI_VDEVICE(MARVELL, 0x6080), chip_608x },
-	{ PCI_VDEVICE(MARVELL, 0x6081), chip_608x },
-
-	{ PCI_VDEVICE(ADAPTEC2, 0x0241), chip_604x },
-
-	/* Adaptec 1430SA */
-	{ PCI_VDEVICE(ADAPTEC2, 0x0243), chip_7042 },
-
-	/* Marvell 7042 support */
-	{ PCI_VDEVICE(MARVELL, 0x7042), chip_7042 },
-
-	/* Highpoint RocketRAID PCIe series */
-	{ PCI_VDEVICE(TTI, 0x2300), chip_7042 },
-	{ PCI_VDEVICE(TTI, 0x2310), chip_7042 },
-
-	{ }			/* terminate list */
-};
-
 static const struct mv_hw_ops mv5xxx_ops = {
 	.phy_errata		= mv5_phy_errata,
 	.enable_leds		= mv5_enable_leds,

@@ -4303,6 +4272,37 @@ static int mv_pci_init_one(struct pci_dev *pdev,
 static int mv_pci_device_resume(struct pci_dev *pdev);
 #endif
 
+static const struct pci_device_id mv_pci_tbl[] = {
+	{ PCI_VDEVICE(MARVELL, 0x5040), chip_504x },
+	{ PCI_VDEVICE(MARVELL, 0x5041), chip_504x },
+	{ PCI_VDEVICE(MARVELL, 0x5080), chip_5080 },
+	{ PCI_VDEVICE(MARVELL, 0x5081), chip_508x },
+	/* RocketRAID 1720/174x have different identifiers */
+	{ PCI_VDEVICE(TTI, 0x1720), chip_6042 },
+	{ PCI_VDEVICE(TTI, 0x1740), chip_6042 },
+	{ PCI_VDEVICE(TTI, 0x1742), chip_6042 },
+
+	{ PCI_VDEVICE(MARVELL, 0x6040), chip_604x },
+	{ PCI_VDEVICE(MARVELL, 0x6041), chip_604x },
+	{ PCI_VDEVICE(MARVELL, 0x6042), chip_6042 },
+	{ PCI_VDEVICE(MARVELL, 0x6080), chip_608x },
+	{ PCI_VDEVICE(MARVELL, 0x6081), chip_608x },
+
+	{ PCI_VDEVICE(ADAPTEC2, 0x0241), chip_604x },
+
+	/* Adaptec 1430SA */
+	{ PCI_VDEVICE(ADAPTEC2, 0x0243), chip_7042 },
+
+	/* Marvell 7042 support */
+	{ PCI_VDEVICE(MARVELL, 0x7042), chip_7042 },
+
+	/* Highpoint RocketRAID PCIe series */
+	{ PCI_VDEVICE(TTI, 0x2300), chip_7042 },
+	{ PCI_VDEVICE(TTI, 0x2310), chip_7042 },
+
+	{ }			/* terminate list */
+};
+
 
 static struct pci_driver mv_pci_driver = {
 	.name			= DRV_NAME,

@@ -4315,6 +4315,7 @@ static struct pci_driver mv_pci_driver = {
 #endif
 
 };
+MODULE_DEVICE_TABLE(pci, mv_pci_tbl);
 
 /**
  *      mv_print_info - Dump key info to kernel log for perusal.

@@ -4487,7 +4488,6 @@ static void __exit mv_exit(void)
 MODULE_AUTHOR("Brett Russ");
 MODULE_DESCRIPTION("SCSI low-level driver for Marvell SATA controllers");
 MODULE_LICENSE("GPL v2");
-MODULE_DEVICE_TABLE(pci, mv_pci_tbl);
 MODULE_VERSION(DRV_VERSION);
 MODULE_ALIAS("platform:" DRV_NAME);
 

Quoted 5 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

Building with W=1 shows a warning for an unused variable when CONFIG_PCI is diabled:

Quoted 8 lines

drivers/ata/sata_mv.c:790:35: error: unused variable 'mv_pci_tbl' [-Werror,-Wunused-const-variable] static const struct pci_device_id mv_pci_tbl[] = {

Move the table into the same block that containsn the pci_driver definition.

Fixes 7bb3c5290ca0 ("sata_mv: Remove PCI dependency")

Signed-off-by Arnd Bergmann <[email protected]>

The patch title is also not describing what the patch does. Are you OK with changing that to:

ata: sata_mv: Fix PCI device ID table declaration warning

?

Quoted 15 lines · Damien Le Moal

Quoted 13 lines · Arnd Bergmann

From: Arnd Bergmann <[email protected]>

Building with W=1 shows a warning for an unused variable when CONFIG_PCI is diabled:

drivers/ata/sata_mv.c:790:35: error: unused variable 'mv_pci_tbl' [-Werror,-Wunused-const-variable] static const struct pci_device_id mv_pci_tbl[] = {

Move the table into the same block that containsn the pci_driver definition.

Fixes 7bb3c5290ca0 ("sata_mv: Remove PCI dependency")

Signed-off-by Arnd Bergmann <[email protected]>

Quoted 6 lines

The patch title is also not describing what the patch does. Are you OK with changing that to:

ata: sata_mv: Fix PCI device ID table declaration warning

?

Yes, please do, thanks!

I had first tried to remove all the #ifdef checks and just rely on dead-code-elimination doing the same when pci_register_driver() is stubbed out and IS_ENABLED(CONFIG_OF) checks turn off the rest. Unfortunately, the include/linux/pci.h interfaces are not all stubbed out here and cause compile-time failures without CONFIG_PCI, so that didn't work out.

Arnd

From: Arnd Bergmann <[email protected]>

Building with W=1 shows a warning for an unused variable when CONFIG_PCI is diabled:

drivers/ata/sata_mv.c:790:35: error: unused variable 'mv_pci_tbl' [-Werror,-Wunused-const-variable] static const struct pci_device_id mv_pci_tbl[] = {

Move the table into the same block that containsn the pci_driver definition.

static struct pci_driver mv_pci_driver = {
.name = DRV_NAME,

From: Arnd Bergmann <[email protected]>

When building with CONFIG_OF disabled but W=1 extra warnings enabled, a couple of driver cause a warning about an unused ID table:

drivers/spi/spi-armada-3700.c:806:34: error: unused variable 'a3700_spi_dt_ids' [-Werror,-Wunused-const-variable] drivers/spi/spi-orion.c:614:34: error: unused variable 'orion_spi_of_match_table' [-Werror,-Wunused-const-variable] drivers/spi/spi-pic32-sqi.c:673:34: error: unused variable 'pic32_sqi_of_ids' [-Werror,-Wunused-const-variable] drivers/spi/spi-pic32.c:850:34: error: unused variable 'pic32_spi_of_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-rockchip.c:1020:34: error: unused variable 'rockchip_spi_dt_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-s3c64xx.c:1642:34: error: unused variable 's3c64xx_spi_dt_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-st-ssc4.c:439:34: error: unused variable 'stm_spi_match' [-Werror,-Wunused-const-variable]

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

From: Arnd Bergmann <[email protected]>

When building with CONFIG_OF disabled but W=1 extra warnings enabled, a couple of driver cause a warning about an unused ID table:

drivers/spi/spi-armada-3700.c:806:34: error: unused variable 'a3700_spi_dt_ids' [-Werror,-Wunused-const-variable] drivers/spi/spi-orion.c:614:34: error: unused variable 'orion_spi_of_match_table' [-Werror,-Wunused-const-variable] drivers/spi/spi-pic32-sqi.c:673:34: error: unused variable 'pic32_sqi_of_ids' [-Werror,-Wunused-const-variable] drivers/spi/spi-pic32.c:850:34: error: unused variable 'pic32_spi_of_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-rockchip.c:1020:34: error: unused variable 'rockchip_spi_dt_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-s3c64xx.c:1642:34: error: unused variable 's3c64xx_spi_dt_match' [-Werror,-Wunused-const-variable] drivers/spi/spi-st-ssc4.c:439:34: error: unused variable 'stm_spi_match' [-Werror,-Wunused-const-variable]

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

Quoted 4 lines · Arnd Bergmann

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

Why are we not fixing of_match_ptr() here, or at least adding the ifdefs in case someone does end up wanting to run without OF?

Quoted 9 lines · Arnd Bergmann

Quoted 5 lines

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

Why are we not fixing of_match_ptr() here, or at least adding the ifdefs in case someone does end up wanting to run without OF?

Fixing of_match_ptr =

include/linux/of.h +0 -0

index a0bedd038a05..d980bccffda0 100644

include/linux/of.h +0 -0

 #define of_match_node(_matches, _node)	NULL
 #endif /* CONFIG_OF */
 

?

Assuming this helps, I agree this would be the better fix.

Wed, Apr 03, 2024 at 11:05:51PM +0200, Uwe Kleine-König kirjoitti:

Quoted 21 lines · Mark Brown

Quoted 9 lines · Arnd Bergmann

Quoted 5 lines

These appear to all be copied from the same original driver, so fix them at the same time by removing the unnecessary of_match_ptr() annotation. As far as I can tell, all these drivers are only actually used on configurations that have CONFIG_OF enabled.

Why are we not fixing of_match_ptr() here, or at least adding the ifdefs in case someone does end up wanting to run without OF?

Fixing of_match_ptr =

include/linux/of.h +0 -0

index a0bedd038a05..d980bccffda0 100644
@@ -890,7 +890,7 @@ static inline const void *of_device_get_match_data(const struct device *dev)
 	return NULL;
 }
 

include/linux/of.h static inline const void *of_device_get_match_data(const struct device *dev) · line 893

Quoted 2 lines

include/linux/of.h +1 -1

-#define of_match_ptr(_ptr)	NULL
+#define of_match_ptr(_ptr)	(0 ? (_ptr) : NULL)

Quoted 5 lines

include/linux/of.h +0 -0

 #define of_match_node(_matches, _node)	NULL
 #endif /* CONFIG_OF */
 

?

Quoted 1 line

Assuming this helps, I agree this would be the better fix.

Why? I mean why do we need to even have this API? It's always good to know which devices are supported by the module even if you have no need in such support or it's not compiled in. One of the reasons why is to be able to google for compatible hardware, for example.