Add the missing check for platform_get_resource and return error
if it fails.
Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
Signed-off-by: Jiasheng Jiang <[email protected]>
---
Changelog:
v1 -> v2:
1. Replace "-ENOMEM" with "-ENODEV".
---
drivers/mfd/intel-lpss-acpi.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/mfd/intel-lpss-acpi.c b/drivers/mfd/intel-lpss-acpi.c
index a143c8dca2d9..212818aef93e 100644
--- a/drivers/mfd/intel-lpss-acpi.c
+++ b/drivers/mfd/intel-lpss-acpi.c
@@ -183,6 +183,9 @@ static int intel_lpss_acpi_probe(struct platform_device *pdev)
return -ENOMEM;
info->mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!info->mem)
+ return -ENODEV;
+
info->irq = platform_get_irq(pdev, 0);
ret = intel_lpss_probe(&pdev->dev, info);
--
2.25.1
On Fri, 09 Jun 2023, Jiasheng Jiang wrote:
> Add the missing check for platform_get_resource and return error
> if it fails.
>
> Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
> Signed-off-by: Jiasheng Jiang <[email protected]>
> ---
> Changelog:
>
> v1 -> v2:
>
> 1. Replace "-ENOMEM" with "-ENODEV".
> ---
> drivers/mfd/intel-lpss-acpi.c | 3 +++
> 1 file changed, 3 insertions(+)
Applied, thanks
--
Lee Jones [李琼斯]
On Fri, Jun 09, 2023 at 09:48:18AM +0800, Jiasheng Jiang wrote:
> Add the missing check for platform_get_resource and return error
> if it fails.
> Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
This does not fix anything and just introduces a duplication code.
I would like this to be reverted. Should I send one?
--
With Best Regards,
Andy Shevchenko
On Fri, 24 Nov 2023, Andy Shevchenko wrote:
> On Fri, Jun 09, 2023 at 09:48:18AM +0800, Jiasheng Jiang wrote:
> > Add the missing check for platform_get_resource and return error
> > if it fails.
>
> > Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
>
> This does not fix anything and just introduces a duplication code.
> I would like this to be reverted. Should I send one?
Checking this value at the source of obtention and providing and earlier
return with arguably a better return value, all at the cost of an
inexpensive pointer comparison to NULL doesn't sound like a terrible
idea.
If you were committed to the idea of removing it, which I suggest you
reconsider, I would insist that you replace it with at least a comment.
--
Lee Jones [李琼斯]
On Mon, Nov 27, 2023 at 08:53:56AM +0000, Lee Jones wrote:
> On Fri, 24 Nov 2023, Andy Shevchenko wrote:
> > On Fri, Jun 09, 2023 at 09:48:18AM +0800, Jiasheng Jiang wrote:
...
> > > Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
> >
> > This does not fix anything and just introduces a duplication code.
> > I would like this to be reverted. Should I send one?
>
> Checking this value at the source of obtention and providing and earlier
> return with arguably a better return value, all at the cost of an
> inexpensive pointer comparison to NULL doesn't sound like a terrible
> idea.
In general, I agree with you, but the cases similar to this. Why?
Because memory resource retrieval and remapping has a lot of helpers,
some of which are enriched with own error handling and messaging.
Yes, we use devm_ioremap_uc(), which doesn't give that (yet?).
However, it will be more work if we, theoretically, switch to
something like devm_ioremap_resource() in the future.
Hence, I would like to have a common code to be in common place
and behave in the same way independently on the glue druver (PCI,
ACPI, etc).
> If you were committed to the idea of removing it, which I suggest you
> reconsider, I would insist that you replace it with at least a comment.
Isn't what I have done in the series I sent last week?
--
With Best Regards,
Andy Shevchenko
On Mon, 27 Nov 2023, Andy Shevchenko wrote:
> On Mon, Nov 27, 2023 at 08:53:56AM +0000, Lee Jones wrote:
> > On Fri, 24 Nov 2023, Andy Shevchenko wrote:
> > > On Fri, Jun 09, 2023 at 09:48:18AM +0800, Jiasheng Jiang wrote:
>
> ...
>
> > > > Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
> > >
> > > This does not fix anything and just introduces a duplication code.
> > > I would like this to be reverted. Should I send one?
> >
> > Checking this value at the source of obtention and providing and earlier
> > return with arguably a better return value, all at the cost of an
> > inexpensive pointer comparison to NULL doesn't sound like a terrible
> > idea.
>
> In general, I agree with you, but the cases similar to this. Why?
> Because memory resource retrieval and remapping has a lot of helpers,
> some of which are enriched with own error handling and messaging.
>
> Yes, we use devm_ioremap_uc(), which doesn't give that (yet?).
> However, it will be more work if we, theoretically, switch to
> something like devm_ioremap_resource() in the future.
>
> Hence, I would like to have a common code to be in common place
> and behave in the same way independently on the glue druver (PCI,
> ACPI, etc).
>
> > If you were committed to the idea of removing it, which I suggest you
> > reconsider, I would insist that you replace it with at least a comment.
>
> Isn't what I have done in the series I sent last week?
You have to send links when you say things like this.
Last week I had 1000 unread upstream mails in this inbox.
--
Lee Jones [李琼斯]