[PATCH v2 1/2] drm/amdgpu: Fix pci platform speed and width
Alex Deucher
alexdeucher at gmail.com
Wed Feb 6 01:13:58 UTC 2019
On Tue, Feb 5, 2019 at 4:38 PM Kasiviswanathan, Harish
<Harish.Kasiviswanathan at amd.com> wrote:
>
> The new Vega series GPU cards have in-built bridges. To get the pcie
> speed and width supported by the platform walk the hierarchy and get the
> slowest link.
>
> Change-Id: I3196d158b0c614cbb5d7a34c793a58cb95322d32
> Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan at amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 58 +++++++++++++++++++++++-------
> 1 file changed, 46 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> index d7dddb9..fcab1fe 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> @@ -3618,6 +3618,38 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
> return r;
> }
>
> +static void amdgpu_device_get_min_pci_speed_width(struct amdgpu_device *adev,
> + enum pci_bus_speed *speed,
> + enum pcie_link_width *width)
> +{
> + struct pci_dev *pdev = adev->pdev;
> + enum pci_bus_speed cur_speed;
> + enum pcie_link_width cur_width;
> +
> + *speed = PCI_SPEED_UNKNOWN;
> + *width = PCIE_LNK_WIDTH_UNKNOWN;
> +
> + while (pdev) {
> + cur_speed = pcie_get_speed_cap(pdev);
> + cur_width = pcie_get_width_cap(pdev);
> +
> + if (cur_speed != PCI_SPEED_UNKNOWN) {
> + if (*speed == PCI_SPEED_UNKNOWN)
> + *speed = cur_speed;
> + else if (cur_speed < *speed)
> + *speed = cur_speed;
> + }
> +
> + if (cur_width != PCIE_LNK_WIDTH_UNKNOWN) {
> + if (*width == PCIE_LNK_WIDTH_UNKNOWN)
> + *width = cur_width;
> + else if (cur_width < *width)
> + *width = cur_width;
> + }
> + pdev = pci_upstream_bridge(pdev);
> + }
> +}
> +
> /**
> * amdgpu_device_get_pcie_info - fence pcie info about the PCIE slot
> *
> @@ -3630,8 +3662,8 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
> static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
> {
> struct pci_dev *pdev;
> - enum pci_bus_speed speed_cap;
> - enum pcie_link_width link_width;
> + enum pci_bus_speed speed_cap, platform_speed_cap;
> + enum pcie_link_width platform_link_width;
>
> if (amdgpu_pcie_gen_cap)
> adev->pm.pcie_gen_mask = amdgpu_pcie_gen_cap;
> @@ -3648,6 +3680,12 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
> return;
> }
>
> + if (adev->pm.pcie_gen_mask && adev->pm.pcie_mlw_mask)
> + return;
I think you can drop this check, but I guess it doesn't hurt. Either
way, series is:
Reviewed-by: Alex Deucher <alexander.deucher at amd.com>
> +
> + amdgpu_device_get_min_pci_speed_width(adev, &platform_speed_cap,
> + &platform_link_width);
> +
> if (adev->pm.pcie_gen_mask == 0) {
> /* asic caps */
> pdev = adev->pdev;
> @@ -3673,22 +3711,20 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
> adev->pm.pcie_gen_mask |= CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1;
> }
> /* platform caps */
> - pdev = adev->ddev->pdev->bus->self;
> - speed_cap = pcie_get_speed_cap(pdev);
> - if (speed_cap == PCI_SPEED_UNKNOWN) {
> + if (platform_speed_cap == PCI_SPEED_UNKNOWN) {
> adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
> } else {
> - if (speed_cap == PCIE_SPEED_16_0GT)
> + if (platform_speed_cap == PCIE_SPEED_16_0GT)
> adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4);
> - else if (speed_cap == PCIE_SPEED_8_0GT)
> + else if (platform_speed_cap == PCIE_SPEED_8_0GT)
> adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3);
> - else if (speed_cap == PCIE_SPEED_5_0GT)
> + else if (platform_speed_cap == PCIE_SPEED_5_0GT)
> adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
> CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
> else
> @@ -3697,12 +3733,10 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
> }
> }
> if (adev->pm.pcie_mlw_mask == 0) {
> - pdev = adev->ddev->pdev->bus->self;
> - link_width = pcie_get_width_cap(pdev);
> - if (link_width == PCIE_LNK_WIDTH_UNKNOWN) {
> + if (platform_link_width == PCIE_LNK_WIDTH_UNKNOWN) {
> adev->pm.pcie_mlw_mask |= AMDGPU_DEFAULT_PCIE_MLW_MASK;
> } else {
> - switch (link_width) {
> + switch (platform_link_width) {
> case PCIE_LNK_X32:
> adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X32 |
> CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 |
> --
> 2.7.4
>
> _______________________________________________
> amd-gfx mailing list
> amd-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
More information about the amd-gfx
mailing list