Date:   Wed, 18 Mar 2020 11:47:58 +0800
From:   Xu Yilun <yilun.xu@...el.com>
To:     Wu Hao <hao.wu@...el.com>
Cc:     mdf@...nel.org, linux-fpga@...r.kernel.org,
        linux-kernel@...r.kernel.org, Luwei Kang <luwei.kang@...el.com>
Subject: Re: [PATCH v2 1/7] fpga: dfl: parse interrupt info for feature
  devices on enumeration

On Tue, Mar 17, 2020 at 05:00:44PM +0800, Wu Hao wrote:
> Hi Yilun
> 
> Some comments inline.
> 
> On Mon, Mar 16, 2020 at 12:16:56PM +0800, Xu Yilun wrote:
> > DFL based FPGA devices could support interrupts for different purposes,
> > but current DFL framework only supports feature device enumeration with
> > given MMIO resources information via common DFL headers. This patch
> > introduces one new API dfl_fpga_enum_info_add_irq for low level bus
> > drivers (e.g. PCIe device driver) to pass its interrupt resources
> > information to DFL framework for enumeration, and also adds interrupt
> > enumeration code in framework to parse and assign interrupt resources
> > for enumerated feature devices and their own sub features.
> > 
> > With this patch, DFL framework enumerates interrupt resources for core
> > features, including PORT Error Reporting, FME (FPGA Management Engine)
> > Error Reporting and also AFU User Interrupts.
> > 
> > Signed-off-by: Luwei Kang <luwei.kang@...el.com>
> > Signed-off-by: Wu Hao <hao.wu@...el.com>
> > Signed-off-by: Xu Yilun <yilun.xu@...el.com>
> > ----
> > v2: early validating irq table for each feature in parse_feature_irq().
> >     Some code improvement and minor fix for Hao's comments.
> > ---
> >  drivers/fpga/dfl.c | 152 +++++++++++++++++++++++++++++++++++++++++++++++++++--
> >  drivers/fpga/dfl.h |  40 ++++++++++++++
> >  2 files changed, 188 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/fpga/dfl.c b/drivers/fpga/dfl.c
> > index 9909948..28e2cd8 100644
> > --- a/drivers/fpga/dfl.c
> > +++ b/drivers/fpga/dfl.c
> > @@ -11,6 +11,7 @@
> >   *   Xiao Guangrong <guangrong.xiao@...ux.intel.com>
> >   */
> >  #include <linux/module.h>
> > +#include <asm/irq.h>
> >  
> >  #include "dfl.h"
> >  
> > @@ -421,6 +422,9 @@ EXPORT_SYMBOL_GPL(dfl_fpga_dev_ops_unregister);
> >   *
> >   * @dev: device to enumerate.
> >   * @cdev: the container device for all feature devices.
> > + * @nr_irqs: number of irqs for all feature devices.
> > + * @irq_table: Linux IRQ numbers for all irqs, indexed by local irq index of
> > + *	       this device.
> >   * @feature_dev: current feature device.
> >   * @ioaddr: header register region address of feature device in enumeration.
> >   * @sub_features: a sub features linked list for feature device in enumeration.
> > @@ -429,6 +433,9 @@ EXPORT_SYMBOL_GPL(dfl_fpga_dev_ops_unregister);
> >  struct build_feature_devs_info {
> >  	struct device *dev;
> >  	struct dfl_fpga_cdev *cdev;
> > +	unsigned int nr_irqs;
> > +	int *irq_table;
> > +
> >  	struct platform_device *feature_dev;
> >  	void __iomem *ioaddr;
> >  	struct list_head sub_features;
> > @@ -442,12 +449,16 @@ struct build_feature_devs_info {
> >   * @mmio_res: mmio resource of this sub feature.
> >   * @ioaddr: mapped base address of mmio resource.
> >   * @node: node in sub_features linked list.
> > + * @irq_base: start of irq index in this sub feature.
> > + * @nr_irqs: number of irqs of this sub feature.
> >   */
> >  struct dfl_feature_info {
> >  	u64 fid;
> >  	struct resource mmio_res;
> >  	void __iomem *ioaddr;
> >  	struct list_head node;
> > +	unsigned int irq_base;
> > +	unsigned int nr_irqs;
> >  };
> >  
> >  static void dfl_fpga_cdev_add_port_dev(struct dfl_fpga_cdev *cdev,
> > @@ -520,6 +531,8 @@ static int build_info_commit_dev(struct build_feature_devs_info *binfo)
> >  	/* fill features and resource information for feature dev */
> >  	list_for_each_entry_safe(finfo, p, &binfo->sub_features, node) {
> >  		struct dfl_feature *feature = &pdata->features[index];
> > +		struct dfl_feature_irq_ctx *ctx;
> > +		int i;
> 
> should be unsigned int?

Yes.

> 
> >  
> >  		/* save resource information for each feature */
> >  		feature->id = finfo->fid;
> > @@ -527,6 +540,20 @@ static int build_info_commit_dev(struct build_feature_devs_info *binfo)
> >  		feature->ioaddr = finfo->ioaddr;
> >  		fdev->resource[index++] = finfo->mmio_res;
> >  
> > +		if (finfo->nr_irqs) {
> > +			ctx = devm_kcalloc(binfo->dev, finfo->nr_irqs,
> > +					   sizeof(*ctx), GFP_KERNEL);
> > +			if (!ctx)
> > +				return -ENOMEM;
> > +
> > +			for (i = 0; i < finfo->nr_irqs; i++)
> > +				ctx[i].irq =
> > +					binfo->irq_table[finfo->irq_base + i];
> > +
> > +			feature->irq_ctx = ctx;
> > +			feature->nr_irqs = finfo->nr_irqs;
> > +		}
> > +
> >  		list_del(&finfo->node);
> >  		kfree(finfo);
> >  	}
> > @@ -648,7 +675,8 @@ static u64 feature_id(void __iomem *start)
> >  static int
> >  create_feature_instance(struct build_feature_devs_info *binfo,
> >  			struct dfl_fpga_enum_dfl *dfl, resource_size_t ofst,
> > -			resource_size_t size, u64 fid)
> > +			resource_size_t size, u64 fid, unsigned int irq_base,
> > +			unsigned int nr_irqs)
> >  {
> >  	struct dfl_feature_info *finfo;
> >  
> > @@ -667,6 +695,8 @@ create_feature_instance(struct build_feature_devs_info *binfo,
> >  	finfo->mmio_res.start = dfl->start + ofst;
> >  	finfo->mmio_res.end = finfo->mmio_res.start + size - 1;
> >  	finfo->mmio_res.flags = IORESOURCE_MEM;
> > +	finfo->irq_base = irq_base;
> > +	finfo->nr_irqs = nr_irqs;
> >  	finfo->ioaddr = dfl->ioaddr + ofst;
> >  
> >  	list_add_tail(&finfo->node, &binfo->sub_features);
> > @@ -684,7 +714,8 @@ static int parse_feature_port_afu(struct build_feature_devs_info *binfo,
> >  
> >  	WARN_ON(!size);
> >  
> > -	return create_feature_instance(binfo, dfl, ofst, size, FEATURE_ID_AFU);
> > +	return create_feature_instance(binfo, dfl, ofst, size, FEATURE_ID_AFU,
> > +				       0, 0);
> >  }
> >  
> >  static int parse_feature_afu(struct build_feature_devs_info *binfo,
> > @@ -724,7 +755,7 @@ static int parse_feature_fiu(struct build_feature_devs_info *binfo,
> >  	if (ret)
> >  		return ret;
> >  
> > -	ret = create_feature_instance(binfo, dfl, ofst, 0, 0);
> > +	ret = create_feature_instance(binfo, dfl, ofst, 0, 0, 0, 0);
> >  	if (ret)
> >  		return ret;
> >  	/*
> > @@ -742,17 +773,86 @@ static int parse_feature_fiu(struct build_feature_devs_info *binfo,
> >  	return ret;
> >  }
> >  
> > +static void parse_feature_irqs(struct build_feature_devs_info *binfo,
> > +			       struct dfl_fpga_enum_dfl *dfl,
> > +			       resource_size_t ofst,
> > +			       unsigned int *irq_base, unsigned int *nr_irqs)
> > +{
> > +	unsigned int i;
> > +	u64 id, v;
> > +	int virq;
> > +
> > +	/*
> > +	 * Ideally DFL framework should only read info from DFL header, but
> > +	 * current version DFL only provides mmio resources information for
> > +	 * each feature in DFL Header, no field for interrupt resources.
> > +	 * Some interrupt resources information are provided by specific
> > +	 * mmio registers of each components(e.g. different private features)
> > +	 * which supports interrupt. So in order to parse and assign irq
> > +	 * resources to different components, DFL framework has to look into
> > +	 * specific capability registers of these core private features.
> > +	 *
> > +	 * Once future DFL version supports generic interrupt resources
> > +	 * information in common DFL headers, some generic interrupt parsing
> > +	 * code could be added. But in order to be compatible to old version
> > +	 * DFL, driver may still fall back to these quirks.
> > +	 */
> > +
> > +	id = feature_id((dfl->ioaddr + ofst));
> > +
> > +	if (id == PORT_FEATURE_ID_UINT) {
> > +		v = readq(dfl->ioaddr + ofst + PORT_UINT_CAP);
> > +		*irq_base = FIELD_GET(PORT_UINT_CAP_FST_VECT, v);
> > +		*nr_irqs = FIELD_GET(PORT_UINT_CAP_INT_NUM, v);
> > +	} else if (id == PORT_FEATURE_ID_ERROR) {
> > +		v = readq(dfl->ioaddr + ofst + PORT_ERROR_CAP);
> > +		*irq_base = FIELD_GET(PORT_ERROR_CAP_INT_VECT, v);
> > +		*nr_irqs = FIELD_GET(PORT_ERROR_CAP_SUPP_INT, v);
> > +	} else if (id == FME_FEATURE_ID_GLOBAL_ERR) {
> > +		v = readq(dfl->ioaddr + ofst + FME_ERROR_CAP);
> > +		*irq_base = FIELD_GET(FME_ERROR_CAP_INT_VECT, v);
> > +		*nr_irqs = FIELD_GET(FME_ERROR_CAP_SUPP_INT, v);
> > +	} else {
> > +		return;
> > +	}
> 
> As you know, most features don't support interrupts, but all of them have to
> go through this whole block, do you think if switch is better in this case?

Yes I can change it to switch style.

> 
> > +
> > +	dev_dbg(binfo->dev, "feature: 0x%llx, nr_irqs: %u, irq_base: %u\n",
> > +		(unsigned long long)id, *nr_irqs, *irq_base);
> > +
> > +	if (*irq_base + *nr_irqs > binfo->nr_irqs)
> > +		goto parse_irq_fail;
> > +
> > +	for (i = 0; i < *nr_irqs; i++) {
> > +		virq = binfo->irq_table[*irq_base + i];
> > +		if (virq < 0 || virq > NR_IRQS)
> > +			goto parse_irq_fail;
> > +	}
> > +
> > +	return;
> > +
> > +parse_irq_fail:
> > +	*irq_base = 0;
> > +	*nr_irqs = 0;
> 
> Please add some comments, if parsing failed, then still try to support features
> without inerrupts capability.
> 
> > +	dev_warn(binfo->dev, "Invalid interrupt number in feature 0x%llx\n",
> > +		 (unsigned long long)id);
> 
> 
> I am thinking if it's possibe to put this interrupt parsing code into
> create_feature_instance, as mmio parsing code for the private feature is
> actually inside the create_feature_instance. If we do it this way, we
> don't need to introduce more parameters to create_feature_instance function.
> 
> How do you think?

I think it's reasonable. I'll try to change it.

> 
> 
> > +}
> > +
> >  static int parse_feature_private(struct build_feature_devs_info *binfo,
> >  				 struct dfl_fpga_enum_dfl *dfl,
> >  				 resource_size_t ofst)
> >  {
> > +	unsigned int irq_base = 0, nr_irqs = 0;
> > +
> >  	if (!binfo->feature_dev) {
> >  		dev_err(binfo->dev, "the private feature %llx does not belong to any AFU.\n",
> >  			(unsigned long long)feature_id(dfl->ioaddr + ofst));
> >  		return -EINVAL;
> >  	}
> >  
> > -	return create_feature_instance(binfo, dfl, ofst, 0, 0);
> > +	parse_feature_irqs(binfo, dfl, ofst, &irq_base, &nr_irqs);
> > +
> > +	return create_feature_instance(binfo, dfl, ofst, 0, 0, irq_base,
> > +				       nr_irqs);
> >  }
> >  
> >  /**
> > @@ -853,6 +953,10 @@ void dfl_fpga_enum_info_free(struct dfl_fpga_enum_info *info)
> >  		devm_kfree(dev, dfl);
> >  	}
> >  
> > +	/* remove irq table */
> > +	if (info->irq_table)
> > +		devm_kfree(dev, info->irq_table);
> > +
> >  	devm_kfree(dev, info);
> >  	put_device(dev);
> >  }
> > @@ -892,6 +996,42 @@ int dfl_fpga_enum_info_add_dfl(struct dfl_fpga_enum_info *info,
> >  }
> >  EXPORT_SYMBOL_GPL(dfl_fpga_enum_info_add_dfl);
> >  
> > +/**
> > + * dfl_fpga_enum_info_add_irq - add irq table to enum info
> > + *
> > + * @info: ptr to dfl_fpga_enum_info
> > + * @nr_irqs: number of irqs of the DFL fpga device to be enumerated.
> > + * @irq_table: Linux IRQ numbers for all irqs, indexed by local irq index of
> > + *	       this device.
> > + *
> > + * One FPGA device may have several interrupts. This function adds irq
> > + * information of the DFL fpga device to enum info for next step enumeration.
> > + * This function should be called before dfl_fpga_feature_devs_enumerate().
> > + * Adding multiply irq tables is not supported so it will fail on a second
> > + * call.
> 
> Actually in current dfl implementation, it allows low layer bus driver to hand
> multiple dfls to dfl framework using one enum_info. So that means we expected
> that only one interrupt table shared by all dfls in the same enum_info. If dfls
> can't share the irq table, or logically they are used to represent different
> fpga devices tree, they need to fill more enum_info. I guess we should add more
> descriptions to tell users these limitations.

OK. I think we need to make clear that all dfls on the device are in one irq
domain, so we provide one irq table in one enum_info.

> 
> Hao
> 
> > + *
> > + * Return: 0 on success, negative error code otherwise.
> > + */
> > +int dfl_fpga_enum_info_add_irq(struct dfl_fpga_enum_info *info,
> > +			       unsigned int nr_irqs, int *irq_table)
> > +{
> > +	if (!nr_irqs)
> > +		return -EINVAL;
> > +
> > +	if (info->irq_table)
> > +		return -EEXIST;
> > +
> > +	info->irq_table = devm_kmemdup(info->dev, irq_table,
> > +				       sizeof(int) * nr_irqs, GFP_KERNEL);
> > +	if (!info->irq_table)
> > +		return -ENOMEM;
> > +
> > +	info->nr_irqs = nr_irqs;
> > +
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL_GPL(dfl_fpga_enum_info_add_irq);
> > +
> >  static int remove_feature_dev(struct device *dev, void *data)
> >  {
> >  	struct platform_device *pdev = to_platform_device(dev);
> > @@ -959,6 +1099,10 @@ dfl_fpga_feature_devs_enumerate(struct dfl_fpga_enum_info *info)
> >  	binfo->dev = info->dev;
> >  	binfo->cdev = cdev;
> >  
> > +	binfo->nr_irqs = info->nr_irqs;
> > +	if (info->nr_irqs)
> > +		binfo->irq_table = info->irq_table;
> > +
> >  	/*
> >  	 * start enumeration for all feature devices based on Device Feature
> >  	 * Lists.
> > diff --git a/drivers/fpga/dfl.h b/drivers/fpga/dfl.h
> > index 4a9a33c..6a498cd 100644
> > --- a/drivers/fpga/dfl.h
> > +++ b/drivers/fpga/dfl.h
> > @@ -112,6 +112,13 @@
> >  #define FME_PORT_OFST_ACC_VF	1
> >  #define FME_PORT_OFST_IMP	BIT_ULL(60)
> >  
> > +/* FME Error Capability Register */
> > +#define FME_ERROR_CAP		0x70
> > +
> > +/* FME Error Capability Register Bitfield */
> > +#define FME_ERROR_CAP_SUPP_INT	BIT_ULL(0)		/* Interrupt Support */
> > +#define FME_ERROR_CAP_INT_VECT	GENMASK_ULL(12, 1)	/* Interrupt vector */
> > +
> >  /* PORT Header Register Set */
> >  #define PORT_HDR_DFH		DFH
> >  #define PORT_HDR_GUID_L		GUID_L
> > @@ -145,6 +152,20 @@
> >  #define PORT_STS_PWR_STATE_AP2	2			/* 90% throttling */
> >  #define PORT_STS_PWR_STATE_AP6	6			/* 100% throttling */
> >  
> > +/* Port Error Capability Register */
> > +#define PORT_ERROR_CAP		0x38
> > +
> > +/* Port Error Capability Register Bitfield */
> > +#define PORT_ERROR_CAP_SUPP_INT	BIT_ULL(0)		/* Interrupt Support */
> > +#define PORT_ERROR_CAP_INT_VECT	GENMASK_ULL(12, 1)	/* Interrupt vector */
> > +
> > +/* Port Uint Capability Register */
> > +#define PORT_UINT_CAP		0x8
> > +
> > +/* Port Uint Capability Register Bitfield */
> > +#define PORT_UINT_CAP_INT_NUM	GENMASK_ULL(11, 0)	/* Interrupts num */
> > +#define PORT_UINT_CAP_FST_VECT	GENMASK_ULL(23, 12)	/* First Vector */
> > +
> >  /**
> >   * struct dfl_fpga_port_ops - port ops
> >   *
> > @@ -189,6 +210,15 @@ struct dfl_feature_driver {
> >  };
> >  
> >  /**
> > + * struct dfl_feature_irq_ctx - dfl private feature interrupt context
> > + *
> > + * @irq: Linux IRQ number of this interrupt.
> > + */
> > +struct dfl_feature_irq_ctx {
> > +	int irq;
> > +};
> > +
> > +/**
> >   * struct dfl_feature - sub feature of the feature devices
> >   *
> >   * @id: sub feature id.
> > @@ -196,12 +226,16 @@ struct dfl_feature_driver {
> >   *		    this index is used to find its mmio resource from the
> >   *		    feature dev (platform device)'s reources.
> >   * @ioaddr: mapped mmio resource address.
> > + * @irq_ctx: interrupt context list.
> > + * @nr_irqs: number of interrupt contexts.
> >   * @ops: ops of this sub feature.
> >   */
> >  struct dfl_feature {
> >  	u64 id;
> >  	int resource_index;
> >  	void __iomem *ioaddr;
> > +	struct dfl_feature_irq_ctx *irq_ctx;
> > +	unsigned int nr_irqs;
> >  	const struct dfl_feature_ops *ops;
> >  };
> >  
> > @@ -388,10 +422,14 @@ static inline u8 dfl_feature_revision(void __iomem *base)
> >   *
> >   * @dev: parent device.
> >   * @dfls: list of device feature lists.
> > + * @nr_irqs: number of irqs for all feature devices.
> > + * @irq_table: Linux IRQ numbers for all irqs, indexed by hw irq numbers.
> >   */
> >  struct dfl_fpga_enum_info {
> >  	struct device *dev;
> >  	struct list_head dfls;
> > +	unsigned int nr_irqs;
> > +	int *irq_table;
> >  };
> >  
> >  /**
> > @@ -415,6 +453,8 @@ struct dfl_fpga_enum_info *dfl_fpga_enum_info_alloc(struct device *dev);
> >  int dfl_fpga_enum_info_add_dfl(struct dfl_fpga_enum_info *info,
> >  			       resource_size_t start, resource_size_t len,
> >  			       void __iomem *ioaddr);
> > +int dfl_fpga_enum_info_add_irq(struct dfl_fpga_enum_info *info,
> > +			       unsigned int nr_irqs, int *irq_table);
> >  void dfl_fpga_enum_info_free(struct dfl_fpga_enum_info *info);
> >  
> >  /**
> > -- 
> > 2.7.4

Powered by blists - more mailing lists