RE: [PATCH v11 01/14] EDAC: Add support for EDAC device feature's control

From: Shiju Jose
Date: Mon Sep 09 2024 - 07:19:09 EST


Thank you for the feedbacks.
Apologies for the delay in replying.

>-----Original Message-----
>From: Borislav Petkov <bp@xxxxxxxxx>
>Sent: 03 September 2024 17:35
>To: Shiju Jose <shiju.jose@xxxxxxxxxx>
>Cc: linux-edac@xxxxxxxxxxxxxxx; linux-cxl@xxxxxxxxxxxxxxx; linux-
>acpi@xxxxxxxxxxxxxxx; linux-mm@xxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
>tony.luck@xxxxxxxxx; rafael@xxxxxxxxxx; lenb@xxxxxxxxxx;
>mchehab@xxxxxxxxxx; dan.j.williams@xxxxxxxxx; dave@xxxxxxxxxxxx; Jonathan
>Cameron <jonathan.cameron@xxxxxxxxxx>; dave.jiang@xxxxxxxxx;
>alison.schofield@xxxxxxxxx; vishal.l.verma@xxxxxxxxx; ira.weiny@xxxxxxxxx;
>david@xxxxxxxxxx; Vilas.Sridharan@xxxxxxx; leo.duran@xxxxxxx;
>Yazen.Ghannam@xxxxxxx; rientjes@xxxxxxxxxx; jiaqiyan@xxxxxxxxxx;
>Jon.Grimm@xxxxxxx; dave.hansen@xxxxxxxxxxxxxxx;
>naoya.horiguchi@xxxxxxx; james.morse@xxxxxxx; jthoughton@xxxxxxxxxx;
>somasundaram.a@xxxxxxx; erdemaktas@xxxxxxxxxx; pgonda@xxxxxxxxxx;
>duenwen@xxxxxxxxxx; mike.malvestuto@xxxxxxxxx; gthelen@xxxxxxxxxx;
>wschwartz@xxxxxxxxxxxxxxxxxxx; dferguson@xxxxxxxxxxxxxxxxxxx;
>wbs@xxxxxxxxxxxxxxxxxxxxxx; nifan.cxl@xxxxxxxxx; jgroves@xxxxxxxxxx;
>vsalve@xxxxxxxxxx; tanxiaofei <tanxiaofei@xxxxxxxxxx>; Zengtao (B)
><prime.zeng@xxxxxxxxxxxxx>; Roberto Sassu <roberto.sassu@xxxxxxxxxx>;
>kangkang.shen@xxxxxxxxxxxxx; wanghuiqiang <wanghuiqiang@xxxxxxxxxx>;
>Linuxarm <linuxarm@xxxxxxxxxx>
>Subject: Re: [PATCH v11 01/14] EDAC: Add support for EDAC device feature's
>control
>
>On Fri, Aug 16, 2024 at 05:42:24PM +0100, shiju.jose@xxxxxxxxxx wrote:
>> From: Shiju Jose <shiju.jose@xxxxxxxxxx>
>>
>> Add generic EDAC device feature's control supports registering
>
>"features"
>
>Check your whole set.
Sure. Modified.
>
>> RAS features supported in the system. Driver exposes feature's control
>> attributes to the userspace in
>
>s/the //
Changed.
>
>> /sys/bus/edac/devices/<dev-name>/<ras-feature>/
>>
>> Co-developed-by: Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx>
>> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@xxxxxxxxxx>
>> Signed-off-by: Shiju Jose <shiju.jose@xxxxxxxxxx>
>> ---
>> drivers/edac/edac_device.c | 178
>+++++++++++++++++++++++++++++++++++++
>> include/linux/edac.h | 60 +++++++++++++
>> 2 files changed, 238 insertions(+)
>>
>> diff --git a/drivers/edac/edac_device.c b/drivers/edac/edac_device.c
>> index 621dc2a5d034..635a41db8b5a 100644
>> --- a/drivers/edac/edac_device.c
>> +++ b/drivers/edac/edac_device.c
>> @@ -570,3 +570,181 @@ void edac_device_handle_ue_count(struct
>edac_device_ctl_info *edac_dev,
>> block ? block->name : "N/A", count, msg); }
>> EXPORT_SYMBOL_GPL(edac_device_handle_ue_count);
>> +
>> +/* EDAC device feature */
>> +static void edac_dev_release(struct device *dev) {
>> + struct edac_dev_feat_ctx *ctx =
>> + container_of(dev, struct edac_dev_feat_ctx, dev);
>
>Ew, no, don't do such silly linebreaks pls.
Changed.
>
>> + kfree(ctx->dev.groups);
>> + kfree(ctx);
>> +}
>> +
>> +const struct device_type edac_dev_type = {
>> + .name = "edac_dev",
>> + .release = edac_dev_release,
>> +};
>> +
>> +static void edac_dev_unreg(void *data) {
>> + device_unregister(data);
>> +}
>> +
>> +/**
>> + * edac_dev_feature_init - Init a ras feature
>
>s/ras/RAS/g
>
>Check your whole set.
Sure. Modified.
>
>> + * @parent: client device.
>> + * @dev_data: pointer to struct edac_dev_data.
>
>I can see it is a pointer. What it is used for?
Updated.
>
>> + * @feat: pointer to struct edac_dev_feature.
>> + * @attr_groups: pointer to attribute group's container.
>> + *
>> + * Returns number of scrub feature's attribute groups on success,
>> + * error otherwise.
>> + */
>> +static int edac_dev_feat_init(struct device *parent,
>> + struct edac_dev_data *dev_data,
>> + const struct edac_dev_feature *ras_feat,
>> + const struct attribute_group **attr_groups) {
>> + int num;
>> +
>> + switch (ras_feat->feat) {
>> + case RAS_FEAT_SCRUB:
>> + dev_data->scrub_ops = ras_feat->scrub_ops;
>> + dev_data->private = ras_feat->scrub_ctx;
>> + return 1;
>> + case RAS_FEAT_ECS:
>> + num = ras_feat->ecs_info.num_media_frus;
>> + dev_data->ecs_ops = ras_feat->ecs_ops;
>> + dev_data->private = ras_feat->ecs_ctx;
>> + return num;
>> + case RAS_FEAT_PPR:
>> + dev_data->ppr_ops = ras_feat->ppr_ops;
>> + dev_data->private = ras_feat->ppr_ctx;
>> + return 1;
>> + default:
>> + return -EINVAL;
>> + }
>> +}
>> +
>> +/**
>> + * edac_dev_register - register device for ras features with edac
>
>s/edac/EDAC/g
>
>Check your whole set.
Modified.
>
>> + * @parent: client device.
>> + * @name: client device's name.
>> + * @private: parent driver's data to store in the context if any.
>> + * @num_features: number of ras features to register.
>> + * @ras_features: list of ras features to register.
>> + *
>> + * Returns 0 on success, error otherwise.
>> + * The new edac_dev_feat_ctx would be freed automatically.
>> + */
>> +int edac_dev_register(struct device *parent, char *name,
>> + void *private, int num_features,
>> + const struct edac_dev_feature *ras_features) {
>> + const struct attribute_group **ras_attr_groups;
>> + struct edac_dev_data *dev_data;
>> + struct edac_dev_feat_ctx *ctx;
>> + int ppr_cnt = 0, ppr_inst = 0;
>> + int attr_gcnt = 0;
>> + int ret, feat;
>> +
>> + if (!parent || !name || !num_features || !ras_features)
>> + return -EINVAL;
>> +
>> + /* Double parse so we can make space for attributes */
>
>Who's "we"?
>
>Please use passive voice in your comments: no "we" or "I", etc.
Fixed.
>
>Personal pronouns are ambiguous in text, especially with so many
>parties/companies/etc developing the kernel so let's avoid them please.
>
>> + for (feat = 0; feat < num_features; feat++) {
>> + switch (ras_features[feat].feat) {
>> + case RAS_FEAT_SCRUB:
>
>Does this need "fallthrough;" or somesuch?
It is a bug, fixed.
>
>> + case RAS_FEAT_PPR:
>> + attr_gcnt++;
>> + ppr_cnt++;
>> + break;
>> + case RAS_FEAT_ECS:
>> + attr_gcnt +=
>ras_features[feat].ecs_info.num_media_frus;
>> + break;
>> + default:
>> + return -EINVAL;
>> + }
>> + }
>> +
>> + ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
>> + if (!ctx)
>> + return -ENOMEM;
>> +
>> + ctx->dev.parent = parent;
>> + ctx->private = private;
>> +
>> + ras_attr_groups = kcalloc(attr_gcnt + 1, sizeof(*ras_attr_groups),
>GFP_KERNEL);
>> + if (!ras_attr_groups) {
>> + ret = -ENOMEM;
>> + goto ctx_free;
>> + }
>> +
>> + if (ppr_cnt) {
>> + ctx->ppr = kcalloc(ppr_cnt, sizeof(*(ctx->ppr)), GFP_KERNEL);
>> + if (!ctx->ppr) {
>> + ret = -ENOMEM;
>> + goto groups_free;
>> + }
>> + }
>> +
>> + attr_gcnt = 0;
>> + for (feat = 0; feat < num_features; feat++, ras_features++) {
>> + switch (ras_features->feat) {
>> + case RAS_FEAT_SCRUB:
>> + if (!ras_features->scrub_ops)
>> + continue;
>> + dev_data = &ctx->scrub;
>> + break;
>> + case RAS_FEAT_ECS:
>> + if (!ras_features->ecs_ops)
>> + continue;
>> + dev_data = &ctx->ecs;
>> + break;
>> + case RAS_FEAT_PPR:
>> + if (!ras_features->ppr_ops)
>> + continue;
>> + dev_data = &ctx->ppr[ppr_inst];
>> + dev_data->instance = ppr_inst;
>> + ppr_inst++;
>> + break;
>> + default:
>> + ret = -EINVAL;
>> + goto data_mem_free;
>> + }
>> + ret = edac_dev_feat_init(parent, dev_data, ras_features,
>> + &ras_attr_groups[attr_gcnt]);
>> + if (ret < 0)
>> + goto data_mem_free;
>> +
>> + attr_gcnt += ret;
>> + }
>
>Newline.
Added newline.
>
>> + ras_attr_groups[attr_gcnt] = NULL;
>> + ctx->dev.bus = edac_get_sysfs_subsys();
>> + ctx->dev.type = &edac_dev_type;
>> + ctx->dev.groups = ras_attr_groups;
>> + dev_set_drvdata(&ctx->dev, ctx);
>
>Ditto.
Added newline.
>
>> + ret = dev_set_name(&ctx->dev, name);
>> + if (ret)
>> + goto data_mem_free;
>> +
>> + ret = device_register(&ctx->dev);
>> + if (ret) {
>> + put_device(&ctx->dev);
>> + goto data_mem_free;
>> + return ret;
>> + }
>> +
>> + return devm_add_action_or_reset(parent, edac_dev_unreg, &ctx->dev);
>> +
>> +data_mem_free:
>> + if (ppr_cnt)
>> + kfree(ctx->ppr);
>> +groups_free:
>> + kfree(ras_attr_groups);
>> +ctx_free:
>> + kfree(ctx);
>> + return ret;
>> +}
>> +EXPORT_SYMBOL_GPL(edac_dev_register);
>> diff --git a/include/linux/edac.h b/include/linux/edac.h index
>> b4ee8961e623..cc96f55ac714 100644
>> --- a/include/linux/edac.h
>> +++ b/include/linux/edac.h
>> @@ -661,4 +661,64 @@ static inline struct dimm_info
>> *edac_get_dimm(struct mem_ctl_info *mci,
>>
>> return mci->dimms[index];
>> }
>> +
>> +/* EDAC device features */
>> +
>> +#define EDAC_FEAT_NAME_LEN 128
>> +
>> +enum edac_dev_feat {
>> + RAS_FEAT_SCRUB,
>> + RAS_FEAT_ECS,
>> + RAS_FEAT_PPR,
>
>What are those? Comments ontop explaining pls.
Added comments.
>
>> + RAS_FEAT_MAX
>> +};
>> +
>> +struct edac_ecs_ex_info {
>> + u16 num_media_frus;
>> +};
>> +
>> +/*
>> + * EDAC device feature information structure */ struct edac_dev_data
>> +{
>> + union {
>> + const struct edac_scrub_ops *scrub_ops;
>> + const struct edac_ecs_ops *ecs_ops;
>> + const struct edac_ppr_ops *ppr_ops;
>> + };
>> + u8 instance;
>> + void *private;
>> +};
>> +
>> +struct device;
>> +
>> +struct edac_dev_feat_ctx {
>> + struct device dev;
>> + void *private;
>> + struct edac_dev_data scrub;
>> + struct edac_dev_data ecs;
>> + struct edac_dev_data *ppr;
>> +};
>> +
>> +struct edac_dev_feature {
>> + enum edac_dev_feat feat;
>
> ft_type;
Sure. Changed.
>
>> + u8 instance;
>> + union {
>> + const struct edac_scrub_ops *scrub_ops;
>> + const struct edac_ecs_ops *ecs_ops;
>> + const struct edac_ppr_ops *ppr_ops;
>> + };
>> + union {
>> + void *scrub_ctx;
>> + void *ecs_ctx;
>> + void *ppr_ctx;
>> + };
>
>Or drop the silly union and simply do
>
> void *ctx;
Modified.
>
>> + union {
>> + struct edac_ecs_ex_info ecs_info;
>> + };
>
>Union with a single member?!
Removed Union.
>
>> +};
>> +
>> +int edac_dev_register(struct device *parent, char *dev_name,
>> + void *parent_pvt_data, int num_features,
>> + const struct edac_dev_feature *ras_features);
>> #endif /* _LINUX_EDAC_H_ */
>> --
>> 2.34.1
>>
>
>--
>Regards/Gruss,
> Boris.
>
>https://people.kernel.org/tglx/notes-about-netiquette

Thanks,
Shiju