[PATCH 00/21] EDAC: Convert to platform remove callback returning void

[PATCH 00/21] EDAC: Convert to platform remove callback returning void

[Uwe Kleine-König]

Hello,

this series converts all platform drivers below drivers/edac to use
.remove_new(). The motivation is to get rid of an integer return code
that is (mostly) ignored by the platform driver core and error prone on
the driver side. However none of the edac drivers suffered from the easy
to make bug, so all drivers are converted in a trivial way.

See commit 5c5a7680e67b ("platform: Provide a remove callback that
returns no value") for an extended explanation and the eventual goal.

The patch for npcm was already sent back in June
(https://lore.kernel.org/linux-edac/20230628071354.665300-1-u.kleine-koenig@pengutronix.de)
but didn't result in enthusiastic review comments and it wasn't picked
up.

There are no interdependencies between the patches. As there are still
quite a few drivers to convert, I'm happy about every patch that makes
it in. So even if there is a merge conflict with one patch until you
apply, please apply the remainder of this series anyhow. I'll come back
to the part that you (maybe) skipped at a later point.

Best regards
Uwe

Uwe Kleine-König (21):
  EDAC/altera: Convert to platform remove callback returning void
  EDAC/armada_xp: Convert to platform remove callback returning void
  EDAC/aspeed: Convert to platform remove callback returning void
  EDAC/bluefield: Convert to platform remove callback returning void
  EDAC/cell: Convert to platform remove callback returning void
  EDAC/cpc925: Convert to platform remove callback returning void
  EDAC/dmc520: Convert to platform remove callback returning void
  EDAC/highbank_l2: Convert to platform remove callback returning void
  EDAC/highbank_mc: Convert to platform remove callback returning void
  EDAC/mpc85xx: Convert to platform remove callback returning void
  EDAC/npcm: Convert to platform remove callback returning void
  EDAC/octeon-l2c: Convert to platform remove callback returning void
  EDAC/octeon-lmc: Convert to platform remove callback returning void
  EDAC/octeon-pc: Convert to platform remove callback returning void
  EDAC/octeon-pci: Convert to platform remove callback returning void
  EDAC/ppc4xx: Convert to platform remove callback returning void
  EDAC/qcom: Convert to platform remove callback returning void
  EDAC/synopsys: Convert to platform remove callback returning void
  EDAC/ti: Convert to platform remove callback returning void
  EDAC/xgene: Convert to platform remove callback returning void
  EDAC/zynqmp: Convert to platform remove callback returning void

 drivers/edac/altera_edac.c      | 12 ++++--------
 drivers/edac/armada_xp_edac.c   | 12 ++++--------
 drivers/edac/aspeed_edac.c      |  6 ++----
 drivers/edac/bluefield_edac.c   |  6 ++----
 drivers/edac/cell_edac.c        |  5 ++---
 drivers/edac/cpc925_edac.c      |  6 ++----
 drivers/edac/dmc520_edac.c      |  6 ++----
 drivers/edac/highbank_l2_edac.c |  5 ++---
 drivers/edac/highbank_mc_edac.c |  5 ++---
 drivers/edac/mpc85xx_edac.c     | 11 ++++-------
 drivers/edac/npcm_edac.c        |  6 ++----
 drivers/edac/octeon_edac-l2c.c  |  6 ++----
 drivers/edac/octeon_edac-lmc.c  |  5 ++---
 drivers/edac/octeon_edac-pc.c   |  5 ++---
 drivers/edac/octeon_edac-pci.c  |  6 ++----
 drivers/edac/ppc4xx_edac.c      |  7 ++-----
 drivers/edac/qcom_edac.c        |  6 ++----
 drivers/edac/synopsys_edac.c    |  6 ++----
 drivers/edac/ti_edac.c          |  6 ++----
 drivers/edac/xgene_edac.c       |  6 ++----
 drivers/edac/zynqmp_edac.c      |  6 ++----
 21 files changed, 48 insertions(+), 91 deletions(-)


base-commit: 0bb80ecc33a8fb5a682236443c1e740d5c917d1d
-- 
2.40.1

[PATCH 12/21] EDAC/octeon-l2c: Convert to platform remove callback returning void

[Uwe Kleine-König]

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.

To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
 drivers/edac/octeon_edac-l2c.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/edac/octeon_edac-l2c.c b/drivers/edac/octeon_edac-l2c.c
index c33059e9b0be..4015eb9af6fe 100644
--- a/drivers/edac/octeon_edac-l2c.c
+++ b/drivers/edac/octeon_edac-l2c.c
@@ -184,19 +184,17 @@ static int octeon_l2c_probe(struct platform_device *pdev)
 	return -ENXIO;
 }
 
-static int octeon_l2c_remove(struct platform_device *pdev)
+static void octeon_l2c_remove(struct platform_device *pdev)
 {
 	struct edac_device_ctl_info *l2c = platform_get_drvdata(pdev);
 
 	edac_device_del_device(&pdev->dev);
 	edac_device_free_ctl_info(l2c);
-
-	return 0;
 }
 
 static struct platform_driver octeon_l2c_driver = {
 	.probe = octeon_l2c_probe,
-	.remove = octeon_l2c_remove,
+	.remove_new = octeon_l2c_remove,
 	.driver = {
 		   .name = "octeon_l2c_edac",
 	}
-- 
2.40.1

[PATCH 13/21] EDAC/octeon-lmc: Convert to platform remove callback returning void

[Uwe Kleine-König]

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.

To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
 drivers/edac/octeon_edac-lmc.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/edac/octeon_edac-lmc.c b/drivers/edac/octeon_edac-lmc.c
index aeb222ca3ed1..18615cbcd9ea 100644
--- a/drivers/edac/octeon_edac-lmc.c
+++ b/drivers/edac/octeon_edac-lmc.c
@@ -302,18 +302,17 @@ static int octeon_lmc_edac_probe(struct platform_device *pdev)
 	return 0;
 }
 
-static int octeon_lmc_edac_remove(struct platform_device *pdev)
+static void octeon_lmc_edac_remove(struct platform_device *pdev)
 {
 	struct mem_ctl_info *mci = platform_get_drvdata(pdev);
 
 	edac_mc_del_mc(&pdev->dev);
 	edac_mc_free(mci);
-	return 0;
 }
 
 static struct platform_driver octeon_lmc_edac_driver = {
 	.probe = octeon_lmc_edac_probe,
-	.remove = octeon_lmc_edac_remove,
+	.remove_new = octeon_lmc_edac_remove,
 	.driver = {
 		   .name = "octeon_lmc_edac",
 	}
-- 
2.40.1

[PATCH 14/21] EDAC/octeon-pc: Convert to platform remove callback returning void

[Uwe Kleine-König]

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.

To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
 drivers/edac/octeon_edac-pc.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/edac/octeon_edac-pc.c b/drivers/edac/octeon_edac-pc.c
index 754eced59c32..ea8a8e337b1e 100644
--- a/drivers/edac/octeon_edac-pc.c
+++ b/drivers/edac/octeon_edac-pc.c
@@ -119,19 +119,18 @@ static int co_cache_error_probe(struct platform_device *pdev)
 	return -ENXIO;
 }
 
-static int co_cache_error_remove(struct platform_device *pdev)
+static void co_cache_error_remove(struct platform_device *pdev)
 {
 	struct co_cache_error *p = platform_get_drvdata(pdev);
 
 	unregister_co_cache_error_notifier(&p->notifier);
 	edac_device_del_device(&pdev->dev);
 	edac_device_free_ctl_info(p->ed);
-	return 0;
 }
 
 static struct platform_driver co_cache_error_driver = {
 	.probe = co_cache_error_probe,
-	.remove = co_cache_error_remove,
+	.remove_new = co_cache_error_remove,
 	.driver = {
 		   .name = "octeon_pc_edac",
 	}
-- 
2.40.1

[PATCH 15/21] EDAC/octeon-pci: Convert to platform remove callback returning void

[Uwe Kleine-König]

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is ignored (apart
from emitting a warning) and this typically results in resource leaks.

To improve here there is a quest to make the remove callback return
void. In the first step of this quest all drivers are converted to
.remove_new(), which already returns void. Eventually after all drivers
are converted, .remove_new() will be renamed to .remove().

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
---
 drivers/edac/octeon_edac-pci.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/edac/octeon_edac-pci.c b/drivers/edac/octeon_edac-pci.c
index 28b238eecefc..108ad9493cfb 100644
--- a/drivers/edac/octeon_edac-pci.c
+++ b/drivers/edac/octeon_edac-pci.c
@@ -87,19 +87,17 @@ static int octeon_pci_probe(struct platform_device *pdev)
 	return res;
 }
 
-static int octeon_pci_remove(struct platform_device *pdev)
+static void octeon_pci_remove(struct platform_device *pdev)
 {
 	struct edac_pci_ctl_info *pci = platform_get_drvdata(pdev);
 
 	edac_pci_del_device(&pdev->dev);
 	edac_pci_free_ctl_info(pci);
-
-	return 0;
 }
 
 static struct platform_driver octeon_pci_driver = {
 	.probe = octeon_pci_probe,
-	.remove = octeon_pci_remove,
+	.remove_new = octeon_pci_remove,
 	.driver = {
 		   .name = "octeon_pci_edac",
 	}
-- 
2.40.1

Re: [PATCH 13/21] EDAC/octeon-lmc: Convert to platform remove callback returning void

[Philippe Mathieu-Daudé]

On 4/10/23 15:12, Uwe Kleine-König wrote:
> The .remove() callback for a platform driver returns an int which makes
> many driver authors wrongly assume it's possible to do error handling by
> returning an error code. However the value returned is ignored (apart
> from emitting a warning) and this typically results in resource leaks.
> 
> To improve here there is a quest to make the remove callback return
> void. In the first step of this quest all drivers are converted to
> .remove_new(), which already returns void. Eventually after all drivers
> are converted, .remove_new() will be renamed to .remove().
> 
> Trivially convert this driver from always returning zero in the remove
> callback to the void returning variant.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>   drivers/edac/octeon_edac-lmc.c | 5 ++---
>   1 file changed, 2 insertions(+), 3 deletions(-)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Re: [PATCH 14/21] EDAC/octeon-pc: Convert to platform remove callback returning void

[Philippe Mathieu-Daudé]

On 4/10/23 15:12, Uwe Kleine-König wrote:
> The .remove() callback for a platform driver returns an int which makes
> many driver authors wrongly assume it's possible to do error handling by
> returning an error code. However the value returned is ignored (apart
> from emitting a warning) and this typically results in resource leaks.
> 
> To improve here there is a quest to make the remove callback return
> void. In the first step of this quest all drivers are converted to
> .remove_new(), which already returns void. Eventually after all drivers
> are converted, .remove_new() will be renamed to .remove().
> 
> Trivially convert this driver from always returning zero in the remove
> callback to the void returning variant.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>   drivers/edac/octeon_edac-pc.c | 5 ++---
>   1 file changed, 2 insertions(+), 3 deletions(-)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Re: [PATCH 15/21] EDAC/octeon-pci: Convert to platform remove callback returning void

[Philippe Mathieu-Daudé]

On 4/10/23 15:12, Uwe Kleine-König wrote:
> The .remove() callback for a platform driver returns an int which makes
> many driver authors wrongly assume it's possible to do error handling by
> returning an error code. However the value returned is ignored (apart
> from emitting a warning) and this typically results in resource leaks.
> 
> To improve here there is a quest to make the remove callback return
> void. In the first step of this quest all drivers are converted to
> .remove_new(), which already returns void. Eventually after all drivers
> are converted, .remove_new() will be renamed to .remove().
> 
> Trivially convert this driver from always returning zero in the remove
> callback to the void returning variant.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>   drivers/edac/octeon_edac-pci.c | 6 ++----
>   1 file changed, 2 insertions(+), 4 deletions(-)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Re: [PATCH 12/21] EDAC/octeon-l2c: Convert to platform remove callback returning void

[Philippe Mathieu-Daudé]

On 4/10/23 15:12, Uwe Kleine-König wrote:
> The .remove() callback for a platform driver returns an int which makes
> many driver authors wrongly assume it's possible to do error handling by
> returning an error code. However the value returned is ignored (apart
> from emitting a warning) and this typically results in resource leaks.
> 
> To improve here there is a quest to make the remove callback return
> void. In the first step of this quest all drivers are converted to
> .remove_new(), which already returns void. Eventually after all drivers
> are converted, .remove_new() will be renamed to .remove().
> 
> Trivially convert this driver from always returning zero in the remove
> callback to the void returning variant.
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>   drivers/edac/octeon_edac-l2c.c | 6 ++----
>   1 file changed, 2 insertions(+), 4 deletions(-)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Re: [PATCH 00/21] EDAC: Convert to platform remove callback returning void

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 1418 bytes --]

Hello Boris, hello Tony,

On Wed, Oct 04, 2023 at 03:12:33PM +0200, Uwe Kleine-König wrote:
> this series converts all platform drivers below drivers/edac to use
> .remove_new(). The motivation is to get rid of an integer return code
> that is (mostly) ignored by the platform driver core and error prone on
> the driver side. However none of the edac drivers suffered from the easy
> to make bug, so all drivers are converted in a trivial way.
> 
> See commit 5c5a7680e67b ("platform: Provide a remove callback that
> returns no value") for an extended explanation and the eventual goal.
> 
> The patch for npcm was already sent back in June
> (https://lore.kernel.org/linux-edac/20230628071354.665300-1-u.kleine-koenig@pengutronix.de)
> but didn't result in enthusiastic review comments and it wasn't picked
> up.
> 
> There are no interdependencies between the patches. As there are still
> quite a few drivers to convert, I'm happy about every patch that makes
> it in. So even if there is a merge conflict with one patch until you
> apply, please apply the remainder of this series anyhow. I'll come back
> to the part that you (maybe) skipped at a later point.

Any news on this series? Would a resend help?

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH 00/21] EDAC: Convert to platform remove callback returning void

[Borislav Petkov]

On Mon, Nov 20, 2023 at 05:20:54PM +0100, Uwe Kleine-König wrote:
> Any news on this series? Would a resend help?

Nah, lemme have a look.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette

Re: [PATCH 00/21] EDAC: Convert to platform remove callback returning void

[Borislav Petkov]

On Wed, Oct 04, 2023 at 03:12:33PM +0200, Uwe Kleine-König wrote:
> Uwe Kleine-König (21):
>   EDAC/altera: Convert to platform remove callback returning void
>   EDAC/armada_xp: Convert to platform remove callback returning void
>   EDAC/aspeed: Convert to platform remove callback returning void
>   EDAC/bluefield: Convert to platform remove callback returning void
>   EDAC/cell: Convert to platform remove callback returning void
>   EDAC/cpc925: Convert to platform remove callback returning void
>   EDAC/dmc520: Convert to platform remove callback returning void
>   EDAC/highbank_l2: Convert to platform remove callback returning void
>   EDAC/highbank_mc: Convert to platform remove callback returning void
>   EDAC/mpc85xx: Convert to platform remove callback returning void
>   EDAC/npcm: Convert to platform remove callback returning void
>   EDAC/octeon-l2c: Convert to platform remove callback returning void
>   EDAC/octeon-lmc: Convert to platform remove callback returning void
>   EDAC/octeon-pc: Convert to platform remove callback returning void
>   EDAC/octeon-pci: Convert to platform remove callback returning void
>   EDAC/ppc4xx: Convert to platform remove callback returning void
>   EDAC/qcom: Convert to platform remove callback returning void
>   EDAC/synopsys: Convert to platform remove callback returning void
>   EDAC/ti: Convert to platform remove callback returning void
>   EDAC/xgene: Convert to platform remove callback returning void
>   EDAC/zynqmp: Convert to platform remove callback returning void

All applied, thanks.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette