Re: [PATCH v3 2/4] edac: Add support for Amazon's Annapurna Labs L1 EDAC

[Hawa, Hanna]

On 9/3/2019 10:24 AM, Robert Richter wrote:
> On 15.07.19 16:24:07, Hanna Hawa wrote:
> > Adds support for Amazon's Annapurna Labs L1 EDAC driver to detect and
> > report L1 errors.
> >
> > Signed-off-by: Hanna Hawa <hhhawa@xxxxxxxxxx>
> > Reviewed-by: James Morse <james.morse@xxxxxxx>
> > ---
> >   MAINTAINERS               |   6 ++
> >   drivers/edac/Kconfig      |   8 +++
> >   drivers/edac/Makefile     |   1 +
> >   drivers/edac/al_l1_edac.c | 156 ++++++++++++++++++++++++++++++++++++++++++++++
> >   4 files changed, 171 insertions(+)
> >   create mode 100644 drivers/edac/al_l1_edac.c
>
> > diff --git a/drivers/edac/al_l1_edac.c b/drivers/edac/al_l1_edac.c
> > new file mode 100644
> > index 0000000..70510ea
> > --- /dev/null
> > +++ b/drivers/edac/al_l1_edac.c
>
> [...]
>
> > +static void al_l1_edac_cpumerrsr(void *arg)
>
> Could this being named to something meaningful, such as
> *_read_status() or so?
>
> > +{
> > +	struct edac_device_ctl_info *edac_dev = arg;
> > +	int cpu, i;
> > +	u32 ramid, repeat, other, fatal;
> > +	u64 val = read_sysreg_s(ARM_CA57_CPUMERRSR_EL1);
> > +	char msg[AL_L1_EDAC_MSG_MAX];
> > +	int space, count;
> > +	char *p;
> > +
> > +	if (!(FIELD_GET(ARM_CA57_CPUMERRSR_VALID, val)))
> > +		return;
>
> [...]
>
> > +static void al_l1_edac_check(struct edac_device_ctl_info *edac_dev)
> > +{
> > +	on_each_cpu(al_l1_edac_cpumerrsr, edac_dev, 1);
> > +}
> > +
> > +static int al_l1_edac_probe(struct platform_device *pdev)
> > +{
> > +	struct edac_device_ctl_info *edac_dev;
> > +	struct device *dev = &pdev->dev;
> > +	int ret;
> > +
> > +	edac_dev = edac_device_alloc_ctl_info(0, (char *)dev_name(dev), 1, "L",
>
> This type cast looks broken. dev_name() is a constant string already.
>
> Other drivers do not use the dynamically generated dev_name() string
> here, instead a fix string such as mod_name or ctl_name could be used.
> edac_device_alloc_ctl_info() later generates a unique instance name
> derived from name + index.

Ack, will fix and use DRV_NAME.

> Regarding the type, this seems to be an API issue of edac_device_
> alloc_ctl_info() that should actually use const char* in its
> interface. So if needed (from what I wrote above it is not) the type
> in the argument list needs to be changed instead.
>
> > +					      1, 1, NULL, 0,
> > +					      edac_device_alloc_index());
> > +	if (IS_ERR(edac_dev))
> > +		return -ENOMEM;
>
> Use the original error code instead.

Actually it return NULL in case of failure, it was changed in v5 to 
check if error/NULL.

> > +
> > +	edac_dev->edac_check = al_l1_edac_check;
> > +	edac_dev->dev = dev;
> > +	edac_dev->mod_name = DRV_NAME;
> > +	edac_dev->dev_name = dev_name(dev);
> > +	edac_dev->ctl_name = "L1 cache";
>
> Should not contain spaces and maybe a bit more specific.

L1_cache_ecc_err? or L1_cache_err?

> > +	platform_set_drvdata(pdev, edac_dev);
> > +
> > +	ret = edac_device_add_device(edac_dev);
> > +	if (ret) {
> > +		dev_err(dev, "Failed to add L1 edac device\n");
>
> Move this printk below to the error path and maybe print the error
> code. You do not cover the -ENOMEM failure.

Ack.

> -Robert
>
> > +		goto err;
> > +	}
> > +
> > +	return 0;
> > +err:
> > +	edac_device_free_ctl_info(edac_dev);
> > +
> > +	return ret;
> > +}