Re: [PATCH v3 4/4] drm/xe/FLR: Support PCIe FLR

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 24/04/24 16:42, Nilawar, Badal wrote:
>
>
> On 24-04-2024 08:42, Aravind Iddamsetty wrote:
>>
>> On 23/04/24 20:34, Nilawar, Badal wrote:
>>>
>>>
>>> On 22-04-2024 12:27, Aravind Iddamsetty wrote:
>>>> PCI subsystem provides callbacks to inform the driver about a request to
>>>> do function level reset by user, initiated by writing to sysfs entry
>>>> /sys/bus/pci/devices/.../reset. This will allow the driver to handle FLR
>>>> without the need to do unbind and rebind as the driver needs to
>>>> reinitialize the device afresh post FLR.
>>>>
>>>> v2:
>>>> 1. separate out gt idle and pci save/restore to a separate patch (Lucas)
>>>> 2. Fixed the warnings seen around xe_guc_submit_stop, xe_guc_puc_fini
>>>>
>>>> v3: declare xe_pci_err_handlers as static(Michal)
>>>>
>>>> Cc: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>
>>>> Cc: Lucas De Marchi <lucas.demarchi@xxxxxxxxx>
>>>> Cc: Michal Wajdeczko <michal.wajdeczko@xxxxxxxxx>
>>>>
>>>> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>
>>>> Signed-off-by: Aravind Iddamsetty <aravind.iddamsetty@xxxxxxxxxxxxxxx>
>>>> ---
>>>>    drivers/gpu/drm/xe/Makefile          |  1 +
>>>>    drivers/gpu/drm/xe/xe_device_types.h |  3 +
>>>>    drivers/gpu/drm/xe/xe_guc_pc.c       |  4 ++
>>>>    drivers/gpu/drm/xe/xe_pci.c          |  9 ++-
>>>>    drivers/gpu/drm/xe/xe_pci.h          |  2 +
>>>>    drivers/gpu/drm/xe/xe_pci_err.c      | 88 ++++++++++++++++++++++++++++
>>>>    drivers/gpu/drm/xe/xe_pci_err.h      | 13 ++++
>>>>    7 files changed, 119 insertions(+), 1 deletion(-)
>>>>    create mode 100644 drivers/gpu/drm/xe/xe_pci_err.c
>>>>    create mode 100644 drivers/gpu/drm/xe/xe_pci_err.h
>>>>
>>>> diff --git a/drivers/gpu/drm/xe/Makefile b/drivers/gpu/drm/xe/Makefile
>>>> index 8bc62bfbc679..693971a1fac0 100644
>>>> --- a/drivers/gpu/drm/xe/Makefile
>>>> +++ b/drivers/gpu/drm/xe/Makefile
>>>> @@ -117,6 +117,7 @@ xe-y += xe_bb.o \
>>>>        xe_module.o \
>>>>        xe_pat.o \
>>>>        xe_pci.o \
>>>> +    xe_pci_err.o \
>>>>        xe_pcode.o \
>>>>        xe_pm.o \
>>>>        xe_preempt_fence.o \
>>>> diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
>>>> index 0a66555229e9..8c749b378a92 100644
>>>> --- a/drivers/gpu/drm/xe/xe_device_types.h
>>>> +++ b/drivers/gpu/drm/xe/xe_device_types.h
>>>> @@ -465,6 +465,9 @@ struct xe_device {
>>>>        /** @pci_state: PCI state of device */
>>>>        struct pci_saved_state *pci_state;
>>>>    +    /** @pci_device_is_reset: device went through PCIe FLR */
>>>> +    bool pci_device_is_reset;
>>>> +
>>>>        /* private: */
>>>>      #if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
>>>> diff --git a/drivers/gpu/drm/xe/xe_guc_pc.c b/drivers/gpu/drm/xe/xe_guc_pc.c
>>>> index 509649d0e65e..efba0fbe2f5c 100644
>>>> --- a/drivers/gpu/drm/xe/xe_guc_pc.c
>>>> +++ b/drivers/gpu/drm/xe/xe_guc_pc.c
>>>> @@ -902,6 +902,10 @@ static void xe_guc_pc_fini(struct drm_device *drm, void *arg)
>>>>            return;
>>>>        }
>>>>    +    /* We already have done this before going through a reset, so skip here */
>>>> +    if (xe->pci_device_is_reset)
>>>> +        return;
>>>> +
>>>>        XE_WARN_ON(xe_force_wake_get(gt_to_fw(pc_to_gt(pc)), XE_FORCEWAKE_ALL));
>>>>        XE_WARN_ON(xe_guc_pc_gucrc_disable(pc));
>>>>        XE_WARN_ON(xe_guc_pc_stop(pc));
>>>> diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
>>>> index a62300990e19..b5a582afc9e7 100644
>>>> --- a/drivers/gpu/drm/xe/xe_pci.c
>>>> +++ b/drivers/gpu/drm/xe/xe_pci.c
>>>> @@ -23,6 +23,7 @@
>>>>    #include "xe_macros.h"
>>>>    #include "xe_mmio.h"
>>>>    #include "xe_module.h"
>>>> +#include "xe_pci_err.h"
>>>>    #include "xe_pci_types.h"
>>>>    #include "xe_pm.h"
>>>>    #include "xe_sriov.h"
>>>> @@ -738,7 +739,7 @@ static void xe_pci_remove(struct pci_dev *pdev)
>>>>        pci_set_drvdata(pdev, NULL);
>>>>    }
>>>>    -static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
>>>> +int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
>>>>    {
>>>>        const struct xe_device_desc *desc = (const void *)ent->driver_data;
>>>>        const struct xe_subplatform_desc *subplatform_desc;
>>>> @@ -986,6 +987,11 @@ static const struct dev_pm_ops xe_pm_ops = {
>>>>    };
>>>>    #endif
>>>>    +static const struct pci_error_handlers xe_pci_err_handlers = {
>>>> +    .reset_prepare = xe_pci_reset_prepare,
>>>> +    .reset_done = xe_pci_reset_done,
>>>> +};
>>>> +
>>>>    static struct pci_driver xe_pci_driver = {
>>>>        .name = DRIVER_NAME,
>>>>        .id_table = pciidlist,
>>>> @@ -995,6 +1001,7 @@ static struct pci_driver xe_pci_driver = {
>>>>    #ifdef CONFIG_PM_SLEEP
>>>>        .driver.pm = &xe_pm_ops,
>>>>    #endif
>>>> +    .err_handler = &xe_pci_err_handlers,
>>>>    };
>>>>      int xe_register_pci_driver(void)
>>>> diff --git a/drivers/gpu/drm/xe/xe_pci.h b/drivers/gpu/drm/xe/xe_pci.h
>>>> index 73b90a430d1f..9faf5380a09e 100644
>>>> --- a/drivers/gpu/drm/xe/xe_pci.h
>>>> +++ b/drivers/gpu/drm/xe/xe_pci.h
>>>> @@ -7,8 +7,10 @@
>>>>    #define _XE_PCI_H_
>>>>      struct pci_dev;
>>>> +struct pci_device_id;
>>>>      int xe_register_pci_driver(void);
>>>>    void xe_unregister_pci_driver(void);
>>>>    void xe_load_pci_state(struct pci_dev *pdev);
>>>> +int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
>>>>    #endif
>>>> diff --git a/drivers/gpu/drm/xe/xe_pci_err.c b/drivers/gpu/drm/xe/xe_pci_err.c
>>>> new file mode 100644
>>>> index 000000000000..5306925ea2fa
>>>> --- /dev/null
>>>> +++ b/drivers/gpu/drm/xe/xe_pci_err.c
>>>> @@ -0,0 +1,88 @@
>>>> +// SPDX-License-Identifier: MIT
>>>> +/*
>>>> + * Copyright © 2024 Intel Corporation
>>>> + */
>>>> +
>>>> +#include <linux/pci.h>
>>>> +#include <drm/drm_drv.h>
>>>> +
>>>> +#include "xe_device.h"
>>>> +#include "xe_gt.h"
>>>> +#include "xe_gt_printk.h"
>>>> +#include "xe_pci.h"
>>>> +#include "xe_pci_err.h"
>>>> +#include "xe_pm.h"
>>>> +#include "xe_uc.h"
>>>> +
>>>> +/**
>>>> + * xe_pci_reset_prepare - Called when user issued a PCIe reset
>>>> + * via /sys/bus/pci/devices/.../reset.
>>>> + * @pdev: PCI device struct
>>>> + */
>>>> +void xe_pci_reset_prepare(struct pci_dev *pdev)
>>>> +{
>>>> +    struct xe_device *xe = pci_get_drvdata(pdev);
>>>> +    struct xe_gt *gt;
>>>> +    int id, err;
>>>> +
>>>> +    pci_warn(pdev, "preparing for PCIe reset\n");
>>>> +
>>>> +    drm_warn(&xe->drm, "removing device access to userspace\n");
>>>> +    drm_dev_unplug(&xe->drm);
>>>> +
>>>> +    xe_pm_runtime_get(xe);
>>>> +    /* idle the GTs */
>>>> +    for_each_gt(gt, xe, id) {
>>>> +        err = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
>>>> +        if (err)
>>>> +            goto reset;
>>>> +        xe_uc_reset_prepare(&gt->uc);
>>>> +        xe_gt_idle(gt);
>>>> +        err = xe_force_wake_put(gt_to_fw(gt), XE_FORCEWAKE_ALL);
>>>> +        XE_WARN_ON(err);
>>>> +    }
>>>> +    xe_pm_runtime_put(xe);
>>> Is xe_save_pci_state() needed here?
>>
>> No, as the state has already been saved post driver probe.
>
> Thanks for clarification. One more doubt I have is in FLR flow what will happen to buffer objects especially mmaped BOs. Will all the BOs get destroyed? I couldn't figure out it in the code.

buffer ummaping is handled in drm_dev_unplug called above.

Thanks,
Aravind.
>
> Badal
>>
>> Regards,
>> Aravind.
>>>
>>> Regards,
>>> Badal Nilawar> +
>>>> +reset:
>>>> +    pci_disable_device(pdev);
>>>> +}
>>>> +
>>>> +/**
>>>> + * xe_pci_reset_done - Called when PCIe reset is done.
>>>> + * @pdev: PCI device struct
>>>> + */
>>>> +void xe_pci_reset_done(struct pci_dev *pdev)
>>>> +{
>>>> +    const struct pci_device_id *ent = pci_match_id(pdev->driver->id_table, pdev);
>>>> +    struct xe_device *xe = pci_get_drvdata(pdev);
>>>> +
>>>> +    dev_info(&pdev->dev,
>>>> +         "device went through PCIe reset, reenabling the device\n");
>>>> +
>>>> +    if (pci_enable_device(pdev)) {
>>>> +        dev_err(&pdev->dev,
>>>> +            "Cannot re-enable PCI device after reset\n");
>>>> +        return;
>>>> +    }
>>>> +    pci_set_master(pdev);
>>>> +    xe_load_pci_state(pdev);
>>>> +
>>>> +    xe->pci_device_is_reset = true;
>>>> +    /*
>>>> +     * We want to completely clean the driver and even destroy
>>>> +     * the xe private data and reinitialize afresh similar to
>>>> +     * probe
>>>> +     */
>>>> +    pdev->driver->remove(pdev);
>>>> +    if (pci_dev_msi_enabled(pdev))
>>>> +        pci_free_irq_vectors(pdev);
>>>> +
>>>> +    devm_drm_dev_release_action(&xe->drm);
>>>> +    pci_disable_device(pdev);
>>>> +
>>>> +    /*
>>>> +     * if this fails the driver might be in a stale state, only option is
>>>> +     * to unbind and rebind
>>>> +     */
>>>> +    xe_pci_probe(pdev, ent);
>>>> +}
>>>> diff --git a/drivers/gpu/drm/xe/xe_pci_err.h b/drivers/gpu/drm/xe/xe_pci_err.h
>>>> new file mode 100644
>>>> index 000000000000..95a4c8ce9cf1
>>>> --- /dev/null
>>>> +++ b/drivers/gpu/drm/xe/xe_pci_err.h
>>>> @@ -0,0 +1,13 @@
>>>> +/* SPDX-License-Identifier: MIT */
>>>> +/*
>>>> + * Copyright © 2024 Intel Corporation
>>>> + */
>>>> +
>>>> +#ifndef _XE_PCI_ERR_H_
>>>> +#define _XE_PCI_ERR_H_
>>>> +
>>>> +struct pci_dev;
>>>> +
>>>> +void xe_pci_reset_prepare(struct pci_dev *pdev);
>>>> +void xe_pci_reset_done(struct pci_dev *pdev);
>>>> +#endif



[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux