Re: [PATCH V2 45/69] ST SPEAr: PCIE gadget suppport

From: Andrew Morton
Date: Tue Oct 19 2010 - 17:49:05 EST


On Fri, 1 Oct 2010 17:26:05 +0530
Viresh KUMAR <viresh.kumar@xxxxxx> wrote:

> From: Pratyush Anand <pratyush.anand@xxxxxx>
>
> This is a configurable gadget. can be configured by sysfs interface. Any
> IP available at PCIE bus can be programmed to be used by host
> controller.It supoorts both INTX and MSI.
> By default, gadget is configured for INTX and SYSRAM1 is mapped to BAR0
> with size 0x1000
>
>
> ...
>
> +static void enable_dbi_access(struct pcie_app_reg *app_reg)

app_reg should have the __iomem tag.

> +{
> + /* Enable DBI access */
> + writel(readl(&app_reg->slv_armisc) | (1 << AXI_OP_DBI_ACCESS_ID),
> + &app_reg->slv_armisc);
> + writel(readl(&app_reg->slv_awmisc) | (1 << AXI_OP_DBI_ACCESS_ID),
> + &app_reg->slv_awmisc);
> +
> +}
> +
> +static void disable_dbi_access(struct pcie_app_reg *app_reg)

ditto

> +{
> + /* disable DBI access */
> + writel(readl(&app_reg->slv_armisc) & ~(1 << AXI_OP_DBI_ACCESS_ID),
> + &app_reg->slv_armisc);
> + writel(readl(&app_reg->slv_awmisc) & ~(1 << AXI_OP_DBI_ACCESS_ID),
> + &app_reg->slv_awmisc);
> +
> +}
> +
> +static void spear_dbi_read_reg(struct spear_pcie_gadget_config *config,
> + int where, int size, u32 *val)
> +{
> + struct pcie_app_reg *app_reg
> + = (struct pcie_app_reg *) config->va_app_base;

ditto

> + u32 va_address;
> +
> + /* Enable DBI access */
> + enable_dbi_access(app_reg);
> +
> + va_address = (u32)config->va_dbi_base + (where & ~0x3);
> +
> + *val = readl(va_address);
> +
> + if (size == 1)
> + *val = (*val >> (8 * (where & 3))) & 0xff;
> + else if (size == 2)
> + *val = (*val >> (8 * (where & 3))) & 0xffff;
> +
> + /* Disable DBI access */
> + disable_dbi_access(app_reg);
> +}
> +
> +static void spear_dbi_write_reg(struct spear_pcie_gadget_config *config,
> + int where, int size, u32 val)
> +{
> + struct pcie_app_reg *app_reg
> + = (struct pcie_app_reg *) config->va_app_base;

etc.

> + u32 va_address;
> +
> + /* Enable DBI access */
> + enable_dbi_access(app_reg);
> +
> + va_address = (u32)config->va_dbi_base + (where & ~0x3);
> +
> + if (size == 4)
> + writel(val, va_address);
> + else if (size == 2)
> + writew(val, va_address + (where & 2));
> + else if (size == 1)
> + writeb(val, va_address + (where & 3));
> +
> + /* Disable DBI access */
> + disable_dbi_access(app_reg);
> +}
> +
>
> ...
>
> +/**

This token is specifically used to introduce a kerneldoc comment, but
this wasn't a kerneldoc comment.

> + * Tell if a device supports a given PCI capability.
> + * Returns the address of the requested capability structure within the
> + * device's PCI configuration space or 0 in case the device does not
> + * support it. Possible values for @cap:
> + *
> + * %PCI_CAP_ID_PM Power Management
> + * %PCI_CAP_ID_AGP Accelerated Graphics Port
> + * %PCI_CAP_ID_VPD Vital Product Data
> + * %PCI_CAP_ID_SLOTID Slot Identification
> + * %PCI_CAP_ID_MSI Message Signalled Interrupts
> + * %PCI_CAP_ID_CHSWP CompactPCI HotSwap
> + * %PCI_CAP_ID_PCIX PCI-X
> + * %PCI_CAP_ID_EXP PCI Express
> + */
>
> ...
>
> +
> +static ssize_t pcie_gadget_show_link(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;

Check all the __iomems;

> + if (readl(&app_reg->app_status_1) & ((u32)1 << XMLH_LINK_UP_ID))
> + return sprintf(buf, "UP");
> + else
> + return sprintf(buf, "DOWN");
> +}
> +
> +static ssize_t pcie_gadget_store_link(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;
> + char link[10];
> +
> + if (sscanf(buf, "%s", link) != 1)

What happens if strlen(buf) >= 10?

> + return -EINVAL;
> +
> + if (!strcmp(link, "UP"))
> + writel(readl(&app_reg->app_ctrl_0) | (1 << APP_LTSSM_ENABLE_ID),
> + &app_reg->app_ctrl_0);
> + else
> + writel(readl(&app_reg->app_ctrl_0)
> + & ~(1 << APP_LTSSM_ENABLE_ID),
> + &app_reg->app_ctrl_0);
> + return count;
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_store_int_type(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + char int_type[10];
> + u32 cap, vector, vec, flags;
> +
> + if (sscanf(buf, "%s", int_type) != 1)

ditto

> + return -EINVAL;
> +
> + if (!strcmp(int_type, "INTA"))
> + spear_dbi_write_reg(config, PCI_INTERRUPT_LINE, 1, 1);
> +
> + else if (!strcmp(int_type, "MSI")) {
> + vector = config->requested_msi;
> + vec = 0;
> + while (vector > 1) {
> + vector /= 2;
> + vec++;
> + }
> + spear_dbi_write_reg(config, PCI_INTERRUPT_LINE, 1, 0);
> + cap = pci_find_own_capability(config, PCI_CAP_ID_MSI);
> + spear_dbi_read_reg(config, cap + PCI_MSI_FLAGS, 1, &flags);
> + flags &= ~PCI_MSI_FLAGS_QMASK;
> + flags |= vec << 1;
> + spear_dbi_write_reg(config, cap + PCI_MSI_FLAGS, 1, flags);
> + }

No checking for unrecognised input?

> + strcpy(config->int_type, int_type);
> +
> + return count;
> +}
> +
> +static DEVICE_ATTR(int_type, S_IWUSR | S_IRUGO, pcie_gadget_show_int_type,
> + pcie_gadget_store_int_type);
> +
> +static ssize_t pcie_gadget_show_no_of_msi(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;

__iomem

> + u32 cap, vector, vec, flags;
> +
> + if ((readl(&app_reg->msg_status) & (1 << CFG_MSI_EN_ID))
> + != (1 << CFG_MSI_EN_ID))
> + vector = 0;
> + else {
> + cap = pci_find_own_capability(config, PCI_CAP_ID_MSI);
> + spear_dbi_read_reg(config, cap + PCI_MSI_FLAGS, 1, &flags);
> + flags &= ~PCI_MSI_FLAGS_QSIZE;
> + vec = flags >> 4;
> + vector = 1;
> + while (vec--)
> + vector *= 2;
> + }
> + config->configured_msi = vector;
> +
> + return sprintf(buf, "%u", vector);
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_store_inta(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;
> + int en;
> +
> + if (sscanf(buf, "%d", &en) != 1)

strict_strtoul() would be better. It will reject input of the form "42foo".

> + return -EINVAL;
> +
> + if (en)
> + writel(readl(&app_reg->app_ctrl_0) | (1 << SYS_INT_ID),
> + &app_reg->app_ctrl_0);
> + else
> + writel(readl(&app_reg->app_ctrl_0) & ~(1 << SYS_INT_ID),
> + &app_reg->app_ctrl_0);
> +
> + return count;
> +}
> +
> +static DEVICE_ATTR(inta, S_IWUSR, NULL, pcie_gadget_store_inta);
> +
> +static ssize_t pcie_gadget_store_send_msi(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;
> + int vector;
> + u32 ven_msi;
> +
> + if (sscanf(buf, "%d", &vector) != 1)

strict_strtoul().

> + return -EINVAL;
> +
> + if (!config->configured_msi)
> + return -EINVAL;
> +
> + if (vector >= config->configured_msi)
> + return -EINVAL;
> +
> + ven_msi = readl(&app_reg->ven_msi_1);
> + ven_msi &= ~VEN_MSI_FUN_NUM_MASK;
> + ven_msi |= 0 << VEN_MSI_FUN_NUM_ID;
> + ven_msi &= ~VEN_MSI_TC_MASK;
> + ven_msi |= 0 << VEN_MSI_TC_ID;
> + ven_msi &= ~VEN_MSI_VECTOR_MASK;
> + ven_msi |= vector << VEN_MSI_VECTOR_ID;
> +
> + /*generating interrupt for msi vector*/
> + ven_msi |= VEN_MSI_REQ_EN;
> + writel(ven_msi, &app_reg->ven_msi_1);
> + /*need to wait till this bit is cleared, it is not cleared
> + * autometically[Bug RTL] TBD*/
> + udelay(1);
> + ven_msi &= ~VEN_MSI_REQ_EN;
> + writel(ven_msi, &app_reg->ven_msi_1);
> +
> + return count;
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_store_vendor_id(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + u32 id;
> +
> + if (sscanf(buf, "%x", &id) != 1)

strict_strtoul can be used here as well?

> + return -EINVAL;
> +
> + spear_dbi_write_reg(config, PCI_VENDOR_ID, 2, id);
> +
> + return count;
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_store_device_id(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + u32 id;
> +
> + if (sscanf(buf, "%x", &id) != 1)

etc.

> + return -EINVAL;
> +
> + spear_dbi_write_reg(config, PCI_DEVICE_ID, 2, id);
> +
> + return count;
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_store_bar0_size(struct device *dev,
> + struct device_attribute *attr, const char *buf, size_t count)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + u32 size, pos, pos1;
> + u32 no_of_bit = 0;
> +
> + if (sscanf(buf, "%x", &size) != 1)

etc.

> + return -EINVAL;
> + /* as per PCIE specs, min bar size supported is 128 bytes. But
> + * our controller supports min as 256*/
> + if (size <= 0x100)
> + size = 0x100;
> + /* max bar size is 1MB*/
> + else if (size >= 0x100000)
> + size = 0x100000;
> + else {
> + pos = 0;
> + pos1 = 0;
> + while (pos < 21) {
> + pos = find_next_bit((unsigned long *)&size, 21, pos);
> + if (pos != 21)
> + pos1 = pos + 1;
> + pos++;
> + no_of_bit++;
> + }
> + if (no_of_bit == 2)
> + pos1--;
> +
> + size = 1 << pos1;
> + }
> + config->bar0_size = size;
> + spear_dbi_write_reg(config, PCIE_BAR0_MASK_REG, 4, size - 1);
> +
> + return count;
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_show_bar0_address(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + struct spear_pcie_gadget_config *config = dev_get_drvdata(dev);
> + struct pcie_app_reg *app_reg =
> + (struct pcie_app_reg *)config->va_app_base;

etc

> + u32 address = readl(&app_reg->pim0_mem_addr_start);
> +
> + return sprintf(buf, "%x", address);
> +}
> +
>
> ...
>
> +static ssize_t pcie_gadget_show_help(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + char text[] = "\t\tlink read->ltssm status\n \
> + link write->arg1 = UP to enable ltsmm DOWN to disable\n \
> + int_type read->type of supported interrupt\n \
> + int_type write->arg1 = interrupt type to be configured and\n \
> + can be INTA, MSI or NO_INT\n \
> + (select MSI only when you have programmed no_of_msi)\n \
> + no_of_msi read->zero if MSI is not enabled by host\n \
> + and positive value is the number of MSI vector granted\n \
> + no_of_msi write->arg1 = number of MSI vector needed\n \
> + inta write->arg1 = 1 to assert INTA and 0 to de-assert\n \
> + send_msi write->arg1 = MSI vector to be send\n \
> + vendor_id read->programmed vendor id (hex)\n\
> + vendor_id write->arg1 = vendor id(hex) to be programmed\n \
> + device_id read->programmed device id(hex)\n \
> + device_id write->arg1 = device id(hex) to be programmed\n \
> + bar0_size read->size of bar0 in hex\n \
> + bar0_size write->arg1= size of bar0 in hex\n \
> + (default bar0 size is 1000 (hex) bytes)\n \
> + bar0_address read->address of bar0 mapped area in hex\n \
> + bar0_address write->arg1 = address of bar0 mapped area in hex\n\
> + (default mapping of bar0 is SYSRAM1(E0800000)\n \
> + (always program bar size before bar address)\n \
> + (kernel might modify bar size and address to align)\n \
> + (read back bar size and address after writing to check)\n \
> + bar0_rw_offset read->offset of bar0 for which bar0_data \n \
> + will return value\n \
> + bar0_rw_offset write->arg1 = offset of bar0 for which\n \
> + bar0_data will write value\n \
> + bar0_data read->data at bar0_rw_offset\n \
> + bar0_data write->arg1 = data to be written at\n \
> + bar0_rw_offset\n";
> +
> + int size = (sizeof(text) < PAGE_SIZE) ? sizeof(text) : PAGE_SIZE;
> +
> + return snprintf(buf, size, "%s", text);
> +}

What the heck is this??

> +static DEVICE_ATTR(help, S_IRUGO, pcie_gadget_show_help, NULL);
> +
> +static struct attribute *pcie_gadget_attributes[] = {
> + &dev_attr_link.attr,
> + &dev_attr_int_type.attr,
> + &dev_attr_no_of_msi.attr,
> + &dev_attr_inta.attr,
> + &dev_attr_send_msi.attr,
> + &dev_attr_vendor_id.attr,
> + &dev_attr_device_id.attr,
> + &dev_attr_bar0_size.attr,
> + &dev_attr_bar0_address.attr,
> + &dev_attr_bar0_rw_offset.attr,
> + &dev_attr_bar0_data.attr,
> + &dev_attr_help.attr,
> + NULL
> +};
> +
>
> ...
>
> +static int __devinit spear_pcie_gadget_probe(struct platform_device *pdev)
> +{
> + struct resource *res0, *res1;
> + struct spear_pcie_gadget_config *config;
> + unsigned int status = 0;
> + int irq;
> + struct clk *clk;
> +
> + /* get resource for application registers*/
> +
> + res0 = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> + if (!res0) {
> + dev_err(&pdev->dev, "no resource defined\n");
> + return -EBUSY;
> + }
> + if (!request_mem_region(res0->start, resource_size(res0),
> + pdev->name)) {
> + dev_err(&pdev->dev, "pcie gadget region already claimed\n");
> + return -EBUSY;
> + }
> + /* get resource for dbi registers*/
> +
> + res1 = platform_get_resource(pdev, IORESOURCE_MEM, 1);
> + if (!res1) {
> + dev_err(&pdev->dev, "no resource defined\n");
> + goto err_rel_res0;
> + }
> + if (!request_mem_region(res1->start, resource_size(res1),
> + pdev->name)) {
> + dev_err(&pdev->dev, "pcie gadget region already claimed\n");
> + goto err_rel_res0;
> + }
> +
> + config = kzalloc(sizeof(*config), GFP_KERNEL);
> + if (!config) {
> + dev_err(&pdev->dev, "out of memory\n");
> + status = -ENOMEM;
> + goto err_rel_res;
> + }
> +
> + config->va_app_base = ioremap(res0->start, resource_size(res0));
> + if (!config->va_app_base) {
> + dev_err(&pdev->dev, "ioremap fail\n");
> + status = -ENOMEM;
> + goto err_kzalloc;
> + }
> +
> + config->base = (void *)res1->start;

Is that __iomem?

> + config->va_dbi_base = ioremap(res1->start, resource_size(res1));
> + if (!config->va_dbi_base) {
> + dev_err(&pdev->dev, "ioremap fail\n");
> + status = -ENOMEM;
> + goto err_iounmap_app;
> + }
> +
> + dev_set_drvdata(&pdev->dev, config);
> +
> + irq = platform_get_irq(pdev, 0);
> + if (irq < 0) {
> + dev_err(&pdev->dev, "no update irq?\n");
> + status = irq;
> + goto err_iounmap;
> + }
> +
> + status = request_irq(irq, spear_pcie_gadget_irq, 0, pdev->name, NULL);
> + if (status) {
> + dev_err(&pdev->dev, "pcie gadget interrupt IRQ%d already \
> + claimed\n", irq);
> + goto err_get_irq;
> + }
> + /* Register sysfs hooks */
> + status = sysfs_create_group(&pdev->dev.kobj, &pcie_gadget_attr_group);
> + if (status)
> + goto err_irq;
> +
> + /* init basic pcie application registers*/
> + /* do not enable clock if it is PCIE0.Ideally , all controller should
> + * have been independent from others with respect to clock. But PCIE1
> + * and 2 depends on PCIE0.So PCIE0 clk is provided during board init.*/
> + if (pdev->id == 1) {
> + /* Ideally CFG Clock should have been also enabled here. But
> + * it is done currently during board init routne*/
> + clk = clk_get_sys("pcie1", NULL);
> + if (!clk) {
> + pr_err("%s:couldn't get clk for pcie1\n", __func__);
> + goto err_irq;
> + }
> + if (clk_enable(clk)) {
> + pr_err("%s:couldn't enable clk for pcie1\n", __func__);
> + goto err_irq;
> + }
> + } else if (pdev->id == 2) {
> + /* Ideally CFG Clock should have been also enabled here. But
> + * it is done currently during board init routne*/
> + clk = clk_get_sys("pcie2", NULL);
> + if (!clk) {
> + pr_err("%s:couldn't get clk for pcie2\n", __func__);
> + goto err_irq;
> + }
> + if (clk_enable(clk)) {
> + pr_err("%s:couldn't enable clk for pcie2\n", __func__);
> + goto err_irq;
> + }
> + }
> + spear13xx_pcie_device_init(config);
> +
> + return 0;
> +err_irq:
> + free_irq(irq, NULL);
> +err_get_irq:
> + dev_set_drvdata(&pdev->dev, NULL);
> +err_iounmap:
> + iounmap(config->va_dbi_base);
> +err_iounmap_app:
> + iounmap(config->va_app_base);
> +err_kzalloc:
> + kfree(config);
> +err_rel_res:
> + release_mem_region(res1->start, resource_size(res1));
> +err_rel_res0:
> + release_mem_region(res0->start, resource_size(res0));
> + return status;
> +}
> +
>
> ...
>

The driver implements a large, complex userspace interface and afaict
that interface wasn't documented anywhere.

But the interface is the most important part of the driver! It should
be documented in a permanent fashion so that reviewers can understand
and review your proposed interface. They will want to do that before
even looking at the code.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/