Re: [PATCH v12 7/9] ACPI: Translate the I/O range of non-MMIO devices before scanning

[John Garry <john.garry@huawei.com>]

On 04/02/2018 07:45, Rafael J. Wysocki wrote:
> On Tue, Jan 23, 2018 at 5:36 PM, John Garry <john.garry@huawei.com> wrote:
>> On some platforms(such as Hip06/Hip07), the legacy ISA/LPC devices access
>> I/O with some special host-local I/O ports known on x86. As their I/O space
>> are not memory mapped like PCI/PCIE MMIO host bridges, this patch is meant
>> to support a new class of I/O host controllers where the local IO ports of
>> the children devices are translated into the Indirect I/O address space.
>>
>> Through the handler attach callback, all the I/O translations are done
>> before starting the enumeration on children devices and the translated
>> addresses are replaced in the children resources.
>

Hi Rafael,

Thanks for checking.

> The changelog is somewhat dry for a patch adding over 300 lines of new
> code and a new file for that matter.

Will add more. Indeed MFD is not even mentioned.

>
>> Signed-off-by: John Garry <john.garry@huawei.com>
>> Signed-off-by: Zhichang Yuan <yuanzhichang@hisilicon.com>
>> Signed-off-by: Gabriele Paoloni <gabriele.paoloni@huawei.com>
>> ---
>>  drivers/acpi/arm64/Makefile          |   1 +
>>  drivers/acpi/arm64/acpi_indirectio.c | 273 +++++++++++++++++++++++++++++++++++
>>  drivers/acpi/internal.h              |   5 +
>>  drivers/acpi/scan.c                  |   1 +
>>  4 files changed, 280 insertions(+)
>>  create mode 100644 drivers/acpi/arm64/acpi_indirectio.c
>>
>> diff --git a/drivers/acpi/arm64/Makefile b/drivers/acpi/arm64/Makefile
>> index 1017def..f4a7f46 100644
>> --- a/drivers/acpi/arm64/Makefile
>> +++ b/drivers/acpi/arm64/Makefile
>> @@ -1,2 +1,3 @@
>>  obj-$(CONFIG_ACPI_IORT)        += iort.o
>>  obj-$(CONFIG_ACPI_GTDT)        += gtdt.o
>> +obj-$(CONFIG_INDIRECT_PIO)     += acpi_indirectio.o
>> diff --git a/drivers/acpi/arm64/acpi_indirectio.c b/drivers/acpi/arm64/acpi_indirectio.c
>> new file mode 100644
>> index 0000000..2649f57
>> --- /dev/null
>> +++ b/drivers/acpi/arm64/acpi_indirectio.c
>> @@ -0,0 +1,273 @@
>
> SPDX license identifier here?

Will fix. I also need to fixup the other new files in the patchset.

>
>> +/*
>> + * ACPI support for indirect-IO bus.
>> + *
>> + * Copyright (C) 2017 HiSilicon Limited, All Rights Reserved.
>> + * Author: Gabriele Paoloni <gabriele.paoloni@huawei.com>
>> + * Author: Zhichang Yuan <yuanzhichang@hisilicon.com>
>> + * Author: John Garry <john.garry@huawei.com>
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>
> And then you can skip the above.
>
> Also I would like to see some description of what's there in this file
> to appear here.

Sure.

>
>> + */
>> +
>> +#include <linux/acpi.h>
>> +#include <linux/logic_pio.h>
>> +#include <linux/mfd/core.h>
>> +#include <linux/platform_device.h>
>> +
>> +ACPI_MODULE_NAME("indirect IO");
>> +
>> +#define ACPI_INDIRECTIO_NAME_LENGTH 255
>> +
>> +#define INDIRECT_IO_INFO(desc) ((unsigned long)&desc)
>> +
>> +struct acpi_indirectio_mfd_cell {
>> +       struct mfd_cell_acpi_match acpi_match;
>> +       char name[ACPI_INDIRECTIO_NAME_LENGTH];
>> +       char pnpid[ACPI_INDIRECTIO_NAME_LENGTH];
>> +};
>> +
>> +struct acpi_indirectio_host_data {
>> +       resource_size_t io_size;
>> +       resource_size_t io_start;
>> +};
>> +
>> +struct acpi_indirectio_device_desc {
>
> Why don't you use a consistent naming convention and call this
> acpi_indirect_io_device_desc (and analogously everywhere above and
> below)?

Right, I can correct the file to have consistent symbol prefixes.

>
>> +       struct acpi_indirectio_host_data pdata; /* device relevant info data */
>> +       int (*pre_setup)(struct acpi_device *adev,
>> +                        struct acpi_indirectio_host_data *pdata);
>> +};
>> +
>> +static int acpi_translate_logicio_res(struct acpi_device *adev,
>> +               struct acpi_device *host, struct resource *resource)
>> +{
>> +       unsigned long sys_port;
>> +       struct device *dev = &adev->dev;
>> +       resource_size_t length = resource->end - resource->start;
>> +
>> +       sys_port = logic_pio_trans_hwaddr(&host->fwnode, resource->start,
>> +                                       length);
>> +
>> +       if (sys_port == -1) {
>
> Would if (sysp_port < 0) not work here?

I don't think so because sys_port is an unsigned value. I should change 
this if statement to use -1UL.

>
>> +               dev_err(dev, "translate bus-addr(0x%llx) fail!\n",
>> +                       resource->start);
>
> That's not a very informative message.  What are users expected to do
> in response to seeing it?

Right, the user generally would not be able to decipher this, so I 
should add some more info on what the fallout will be.

>
>> +               return -EFAULT;
>> +       }
>> +
>> +       resource->start = sys_port;
>> +       resource->end = sys_port + length;
>> +
>> +       return 0;
>> +}
>> +
>> +/*
>> + * update/set the current I/O resource of the designated device node.
>> + * after this calling, the enumeration can be started as the I/O resource
>> + * had been translated to logicial I/O from bus-local I/O.
>> + *
>> + * @child: the device node to be updated the I/O resource;
>> + * @hostdev: the device node where 'adev' is attached, which can be not
>> + *  the parent of 'adev';
>> + * @res: double pointer to be set to the address of the updated resources
>> + * @num_res: address of the variable to contain the number of updated resources
>> + *
>> + * return 0 when successful, negative is for failure.
>> + */
>
> The above should be a proper kerneldoc comment.
>
>> +int acpi_indirectio_set_logicio_res(struct device *child,

This should actually be static, so I think we can avoid the kerneldoc 
comment?

>> +                                        struct device *hostdev,
>> +                                        const struct resource **res,
>> +                                        int *num_res)
>> +{
>> +       struct acpi_device *adev;
>> +       struct acpi_device *host;
>> +       struct resource_entry *rentry;
>> +       LIST_HEAD(resource_list);
>> +       struct resource *resources = NULL;
>> +       int count;
>> +       int i;
>> +       int ret = -EIO;
>> +
>> +       if (!child || !hostdev)
>> +               return -EINVAL;
>> +
>> +       host = to_acpi_device(hostdev);
>> +       adev = to_acpi_device(child);
>> +
>> +       /* check the device state */
>> +       if (!adev->status.present) {
>> +               dev_info(child, "ACPI: device is not present!\n");
>> +               return 0;
>> +       }
>> +       /* whether the child had been enumerated? */
>> +       if (acpi_device_enumerated(adev)) {
>> +               dev_info(child, "ACPI: had been enumerated!\n");
>> +               return 0;
>> +       }
>> +
>> +       count = acpi_dev_get_resources(adev, &resource_list, NULL, NULL);
>> +       if (count <= 0) {
>> +               dev_err(&adev->dev, "failed to get ACPI resources\n");
>> +               return count ? count : -EIO;
>> +       }
>> +
>> +       resources = kcalloc(count, sizeof(struct resource), GFP_KERNEL);
>> +       if (!resources) {
>> +               acpi_dev_free_resource_list(&resource_list);
>> +               return -ENOMEM;
>> +       }
>> +       count = 0;
>> +       list_for_each_entry(rentry, &resource_list, node)
>> +               resources[count++] = *rentry->res;
>> +
>> +       acpi_dev_free_resource_list(&resource_list);
>> +
>> +       /* translate the I/O resources */
>> +       for (i = 0; i < count; i++) {
>> +               if (resources[i].flags & IORESOURCE_IO) {
>> +                       ret = acpi_translate_logicio_res(adev, host,
>> +                                                       &resources[i]);
>
> You don't need to break this line as far as I'm concerned.

This is just to keep checkpatch happy. I could move the complete 
function call to a single line. And also shortening some symbols will help.

>
>> +                       if (ret) {
>> +                               kfree(resources);
>> +                               dev_err(child,
>> +                                       "Translate I/O range failed (%d)!\n",
>> +                                       ret);
>
> And this too.

Again, just appeasing checkpatch

>
>> +                               return ret;
>> +                       }
>> +               }
>> +       }
>> +       *res = resources;
>> +       *num_res = count;
>> +
>> +       return ret;
>> +}
>> +
>> +int
>
> Don't break the line here, please.

Alright, will do. I thought that this was an accepted style when the 
arguments list is quite long.

>
>> +acpi_indirectio_pre_setup(struct acpi_device *adev,
>> +                         struct acpi_indirectio_host_data *pdata)
>
> As a non-static function, this requires a kerneldoc comment.

In fact this should be static. I should have fixed these already.

>
> In particular, the comment should describe who and when is expected to
> call this function (and which also applies to the comment preceding
> the previous function).
>
> Apparently, they both need to be called before the initial namespace
> scan for the acpi_indirectio_attach() below to work, right?
>

Right, this needs to be prior to device enumeration in the ACPI namespace.

>> +{
>> +       struct platform_device *pdev;
>> +       struct mfd_cell *mfd_cells;
>> +       struct logic_pio_hwaddr *range;
>> +       struct acpi_device *child;
>> +       struct acpi_indirectio_mfd_cell *acpi_indirectio_mfd_cells;
>> +       int size, ret, count = 0, cell_num = 0;
>> +
>> +       range = kzalloc(sizeof(*range), GFP_KERNEL);
>> +       if (!range)
>> +               return -ENOMEM;
>> +       range->fwnode = &adev->fwnode;
>> +       range->flags = PIO_INDIRECT;
>> +       range->size = pdata->io_size;
>> +       range->hw_start = pdata->io_start;
>> +
>> +       ret = logic_pio_register_range(range);
>> +       if (ret)
>> +               goto free_range;
>> +
>> +       list_for_each_entry(child, &adev->children, node)
>> +               cell_num++;
>> +
>> +       /* allocate the mfd cells */
>> +       size = sizeof(*mfd_cells) + sizeof(*acpi_indirectio_mfd_cells);
>> +       mfd_cells = kcalloc(cell_num, size, GFP_KERNEL);
>> +       if (!mfd_cells) {
>> +               ret = -ENOMEM;
>> +               goto free_range;
>> +       }
>> +
>> +       acpi_indirectio_mfd_cells = (struct acpi_indirectio_mfd_cell *)
>> +                                       &mfd_cells[cell_num];
>> +       /* Only consider the children of the host */
>> +       list_for_each_entry(child, &adev->children, node) {
>> +               struct mfd_cell *mfd_cell = &mfd_cells[count];
>> +               struct acpi_indirectio_mfd_cell *acpi_indirectio_mfd_cell =
>> +                                       &acpi_indirectio_mfd_cells[count];
>> +               const struct mfd_cell_acpi_match *acpi_match =
>> +                                       &acpi_indirectio_mfd_cell->acpi_match;
>> +               char *name = &acpi_indirectio_mfd_cell[count].name[0];
>> +               char *pnpid = &acpi_indirectio_mfd_cell[count].pnpid[0];
>> +               struct mfd_cell_acpi_match match = {
>> +                               .pnpid = pnpid,
>> +               };
>> +
>> +               snprintf(name, ACPI_INDIRECTIO_NAME_LENGTH, "indirect-io-%s",
>> +                        acpi_device_hid(child));
>> +               snprintf(pnpid, ACPI_INDIRECTIO_NAME_LENGTH, "%s",
>> +                        acpi_device_hid(child));
>> +
>> +               memcpy((void *)acpi_match, (void *)&match, sizeof(*acpi_match));
>> +               mfd_cell->name = name;
>> +               mfd_cell->acpi_match = acpi_match;
>> +
>> +               ret =
>> +               acpi_indirectio_set_logicio_res(&child->dev,
>> +                                               &adev->dev,
>> +                                               &mfd_cell->resources,
>> +                                               &mfd_cell->num_resources);
>> +               if (ret) {
>> +                       dev_err(&child->dev, "set resource failed (%d)\n", ret);
>> +                       goto free_mfd_res;
>> +               }
>> +               count++;
>> +       }
>> +
>> +       pdev = acpi_create_platform_device(adev, NULL);
>> +       if (IS_ERR_OR_NULL(pdev)) {
>> +               dev_err(&adev->dev, "Create platform device for host failed!\n");
>> +               ret = PTR_ERR(pdev);
>> +               goto free_mfd_res;
>> +       }
>> +       acpi_device_set_enumerated(adev);
>> +
>> +       ret = mfd_add_devices(&pdev->dev, PLATFORM_DEVID_NONE,
>> +                       mfd_cells, cell_num, NULL, 0, NULL);
>> +       if (ret) {
>> +               dev_err(&pdev->dev, "failed to add mfd cells (%d)\n", ret);
>> +               goto free_mfd_res;
>> +       }
>> +
>> +       return ret;
>> +
>> +free_mfd_res:
>> +       while (cell_num--)
>> +               kfree(mfd_cells[cell_num].resources);
>> +       kfree(mfd_cells);
>> +free_range:
>> +       kfree(range);
>> +
>> +       return ret;
>> +}
>> +
>> +/* All the host devices which apply indirect-IO can be listed here. */
>> +static const struct acpi_device_id acpi_indirect_host_id[] = {
>> +       {""},
>> +};
>> +
>> +static int acpi_indirectio_attach(struct acpi_device *adev,
>> +                               const struct acpi_device_id *id)
>> +{
>> +       struct acpi_indirectio_device_desc *hostdata;
>> +       int ret;
>> +
>> +       hostdata = (struct acpi_indirectio_device_desc *)id->driver_data;
>> +       if (!hostdata || !hostdata->pre_setup)
>> +               return -EINVAL;
>> +
>> +       ret = hostdata->pre_setup(adev, &hostdata->pdata);
>> +
>> +       if (ret < 0)
>> +               return ret;
>> +
>> +       return 1;
>> +}
>> +
>> +static struct acpi_scan_handler acpi_indirect_handler = {
>
> acpi_indirect_io_handler?

sure

>
>> +       .ids = acpi_indirect_host_id,
>> +       .attach = acpi_indirectio_attach,
>> +};
>> +
>> +void __init acpi_indirectio_scan_init(void)
>> +{
>> +       acpi_scan_add_handler(&acpi_indirect_handler);
>> +}
>> diff --git a/drivers/acpi/internal.h b/drivers/acpi/internal.h
>> index 1d0a501..d6b1a95 100644
>> --- a/drivers/acpi/internal.h
>> +++ b/drivers/acpi/internal.h
>> @@ -31,6 +31,11 @@
>>  void acpi_platform_init(void);
>>  void acpi_pnp_init(void);
>>  void acpi_int340x_thermal_init(void);
>> +#ifdef CONFIG_INDIRECT_PIO
>> +void acpi_indirectio_scan_init(void);
>> +#else
>> +static inline void acpi_indirectio_scan_init(void) {}
>> +#endif
>>  #ifdef CONFIG_ARM_AMBA
>>  void acpi_amba_init(void);
>>  #else
>> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
>> index b0fe527..8ea6f69 100644
>> --- a/drivers/acpi/scan.c
>> +++ b/drivers/acpi/scan.c
>> @@ -2141,6 +2141,7 @@ int __init acpi_scan_init(void)
>>         acpi_amba_init();
>>         acpi_watchdog_init();
>>         acpi_init_lpit();
>> +       acpi_indirectio_scan_init();
>>
>>         acpi_scan_add_handler(&generic_device_handler);
>>
>> --
>> 1.9.1
>>

Thank you,
John