drm/msm/dpu: fix error check return value of irq_of_parse_and_map()

The irq_of_parse_and_map() function returns 0 on failure, and does not
return a negative value anyhow, so never enter this conditional branch.

Fixes: 25fdd5933e4c ("drm/msm: Add SDM845 DPU support")
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Lv Ruyi <lv.ruyi@zte.com.cn>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/483291/
Link: https://lore.kernel.org/r/20220425090947.3498897-1-lv.ruyi@zte.com.cn
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: Issue MDSS reset during initialization

It's typical for the bootloader to bring up the display for showing a
boot splash or efi framebuffer. But in some cases the kernel driver ends
up only partially configuring (in particular) the DPU, which might
result in e.g. that two different data paths attempts to push data to
the interface - with resulting graphical artifacts.

Naturally the end goal would be to inherit the bootloader's
configuration and provide the user with a glitch free handover from the
boot configuration to a running DPU.

But as implementing seamless transition from the bootloader
configuration to the running OS will be a considerable effort, start by
simply resetting the entire MDSS to its power-on state, to avoid the
partial configuration.

Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482796/
Link: https://lore.kernel.org/r/20220421041550.643964-2-bjorn.andersson@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

dt-bindings: display: msm: Add optional resets

Add an optional reference to the MDSS_CORE reset, which when specified
can be used by the implementation to reset the hardware blocks.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482794/
Link: https://lore.kernel.org/r/20220421041550.643964-1-bjorn.andersson@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: add inline rotation support for sc7280

- Some DPU versions support inline rot90. It is supported only for
limited amount of UBWC formats.
- There are two versions of inline rotators, v1 (present on sm8250 and
sm7250) and v2 (sc7280). These versions differ in the list of supported
formats and in the scaler possibilities.

Co-developed-by: Kalyan Thota <quic_kalyant@quicinc.com>
Signed-off-by: Kalyan Thota <quic_kalyant@quicinc.com>
Signed-off-by: Vinod Polimera <quic_vpolimer@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/481672/
Link: https://lore.kernel.org/r/1649695021-19132-3-git-send-email-quic_vpolimer@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: add inline function to validate format support

Check if the dpu format is supported or not using dpu_find_format.

Co-developed-by: Kalyan Thota <quic_kalyant@quicinc.com>
Signed-off-by: Kalyan Thota <quic_kalyant@quicinc.com>
Signed-off-by: Vinod Polimera <quic_vpolimer@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/481670/
Link: https://lore.kernel.org/r/1649695021-19132-2-git-send-email-quic_vpolimer@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: Bind pingpong block to intf on active ctls in cmd encoder

As per the specification of DPU_CTL_ACTIVE_CFG the configuration of
active blocks should be proactively specified, and the pingpong block is
no different.

The downstream display driver [1] confirms this by also calling
bind_pingpong_blk on CTL_ACTIVE_CFG.  Note that this else-if is always
entered, as setup_intf_cfg - unlike this mainline dpu driver that
combines both behind the same function pointer - is left NULL in favour
of using setup_intf_cfg_v1 when CTL_ACTIVE_CFG is set.

This solves continuous timeouts on at least the Qualcomm sm6125 SoC:

    [drm:dpu_encoder_frame_done_timeout:2091] [dpu error]enc31 frame done timeout
    [drm:_dpu_encoder_phys_cmd_handle_ppdone_timeout.isra.0] *ERROR* id:31 pp:0 kickoff timeout 0 cnt 1 koff_cnt 1
    [drm:dpu_encoder_phys_cmd_prepare_for_kickoff] *ERROR* failed wait_for_idle: id:31 ret:-110 pp:0

In the same way this pingpong block should also be unbound followed by
an interface flush when the encoder is disabled, according to the
downstream display driver [2].

[1]: https://source.codeaurora.org/quic/la/platform/vendor/opensource/display-drivers/tree/msm/sde/sde_encoder_phys_cmd.c?h=LA.UM.9.16.r1-08500-MANNAR.0#n167
[2]: https://source.codeaurora.org/quic/la/platform/vendor/opensource/display-drivers/tree/msm/sde/sde_encoder.c?h=LA.UM.9.16.r1-08500-MANNAR.0#n2986

Signed-off-by: Marijn Suijten <marijn.suijten@somainline.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/475874/
Link: https://lore.kernel.org/r/20220223114011.219044-1-marijn.suijten@somainline.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: drop INTF_EDP from interface type conditions

To remove possible confusion between (old) INTF_EDP and newer INTF_DP,
stop using INTF_EDP in DPU's code. Until the 8x74/8x84 SoCs are
supported by DPU driver, there is no point in using INTF_EDP.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/475560/
Link: https://lore.kernel.org/r/20220222062246.242577-5-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: drop obsolete INTF_EDP comment

DPU driver never supported INTF_EDP, so let's drop the obsolete comment.
If at some point 8x74/8x84's INTF_EDP is ported to DPU driver,
corresponding handling will have to be ported too. Until that time, the
comment serves no purpose.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/475557/
Link: https://lore.kernel.org/r/20220222062246.242577-4-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: drop INTF_TYPE_MAX symbol

This enum value does not correspond to any of actual interface types,
it's not used by the driver, and the value of INTF_WB is greater than
INTF_TYPE_MAX. Thus this symbol serves no purpose and can be removed.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/475556/
Link: https://lore.kernel.org/r/20220222062246.242577-3-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: document INTF_EDP/INTF_DP difference

Based on the discussions on the mailing list, document enum
dpu_intf_type and it's controversial fields: INTF_DP and INTF_EDP.

INTF_EDP is used for older eDP interface found on msm8x74/msm8x84
INTF_DP is used for both eDP and DP interfaces handled by the msm/dp
driver. The DPU driver does not make a difference between them.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/475555/
Link: https://lore.kernel.org/r/20220222062246.242577-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: pass irq to dpu_encoder_helper_wait_for_irq()

Pass IRQ number directly rather than passing an index in the dpu_encoder's irq table.

Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/474703/
Link: https://lore.kernel.org/r/20220217043148.480898-7-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: remove struct dpu_encoder_irq

Remove additional indirection: specify IRQ callbacks and IRQ indices
directly rather than through the pointer in the irq structure. For each
IRQ we have a constant IRQ callback. This change simplifies code review
as the reader no longer needs to remember which function is called.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/474700/
Link: https://lore.kernel.org/r/20220217043148.480898-6-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: get rid of dpu_encoder_helper_(un)register_irq

Get rid of dpu_encoder_helper_register_irq/unregister_irq helpers, call
dpu_core_register/unregister_callback directly, without surrounding them
with helpers.

Reviewed-by: Abhinav Kumar <abhinavk@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/474698/
Link: https://lore.kernel.org/r/20220217043148.480898-5-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: allow just single IRQ callback

DPU interrupts code allows multiple callbacks per interrut. In reality
none of the interrupts is shared between blocks (and will probably never
be). Drop support for registering multiple callbacks per interrupt to
simplify interrupt handling code.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/474701/
Link: https://lore.kernel.org/r/20220217043148.480898-4-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: remove always-true argument of dpu_core_irq_read()

The argument clear of the function dpu_core_irq_read() is always true.
Remove it.

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/474697/
Link: https://lore.kernel.org/r/20220217043148.480898-3-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: remove extra wrappers around dpu_core_irq

Remove extra dpu_irq_* wrappers from dpu_kms.c, merge them directly into
dpu_core_irq_* functions.

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/474696/
Link: https://lore.kernel.org/r/20220217043148.480898-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: change msm_sched_ops from global to static

Smatch reports this issue
msm_ringbuffer.c:43:36: warning: symbol 'msm_sched_ops' was not declared. Should it be static?

msm_sched_ops is only used in msm_ringbuffer.c so change its
storage-class specifier to static.

Signed-off-by: Tom Rix <trix@redhat.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482883/
Link: https://lore.kernel.org/r/20220421131507.1557667-1-trix@redhat.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: remove unused hotplug and edid macros from msm_drv.h

Remove unused MSM_DISPLAY_CAP_HOT_PLUG and MSM_DISPLAY_CAP_EDID
macros from msm_drv.h.

Even if we need these, there are drm equivalent ones present.

Reported-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/482260/
Link: https://lore.kernel.org/r/1650049782-8421-1-git-send-email-quic_abhinavk@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dsi: Add support for DSC configuration

When DSC is enabled, we need to configure DSI registers accordingly and
configure the respective stream compression registers.

Add support to calculate the register setting based on DSC params and
timing information and configure these registers.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480934/
Link: https://lore.kernel.org/r/20220406094031.1027376-15-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: Update generated headers

Update headers from mesa commit:

  commit 28ae397be111c37c6ced397e12d453a7695701bd
  Author: Vinod Koul <vkoul@kernel.org>
  Date:   Fri Apr 1 16:53:04 2022 +0530

      freedreno/registers: update dsi registers to support dsc

      Display Stream compression (DSC) compresses the display stream in
      host which is later decoded by panel. This requires addition of 3 new
      DSI registers to support DSC over DSI.

Signed-off-by: Vinod Koul <vkoul@kernel.org>
      Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/14967>

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480932/
Link: https://lore.kernel.org/r/20220406094031.1027376-14-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dsi: add mode valid callback for dsi_mgr

Add a mode valid callback for dsi_mgr for checking mode being valid in
case of DSC. For DSC the height and width needs to be multiple of slice,
so we check that here

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480930/
Link: https://lore.kernel.org/r/20220406094031.1027376-13-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add DSC support in RM

This add the bits in RM to enable the DSC blocks

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480928/
Link: https://lore.kernel.org/r/20220406094031.1027376-12-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add support for DSC in topology

For DSC to work we typically need a 2,2,1 configuration. This should
suffice for resolutions up to 4k. For more resolutions like 8k this won't
work.

Also, it is better to use 2 LMs and DSC instances as half width results
in lesser power consumption as compared to single LM, DSC at full width.

The panel has been tested only with 2,2,1 configuration, so for
now we blindly create 2,2,1 topology when DSC is enabled

Co-developed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480925/
Link: https://lore.kernel.org/r/20220406094031.1027376-11-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: Add missing num_dspp field documentation

Somehow documentation for num_dspp was missed, so add that

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480924/
Link: https://lore.kernel.org/r/20220406094031.1027376-10-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: don't use merge_3d if DSC merge topology is used

DPU supports different topologies for the case when multiple INTFs are
being driven by the single phys_enc. The driver defaults to using 3DMux
in such cases. Don't use it if DSC merge is used instead.

Suggested-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/480922/
Link: https://lore.kernel.org/r/20220406094031.1027376-9-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add support for DSC in encoder

We need to configure the encoder for DSC configuration and calculate DSC
parameters for the given timing so this patch adds that support by
adding dpu_encoder_prep_dsc() which is invoked when DSC is enabled.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480920/
Link: https://lore.kernel.org/r/20220406094031.1027376-8-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add DSC support in hw_ctl

Later gens of hardware have DSC bits moved to hw_ctl, so configure these
bits so that DSC would work there as well

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480918/
Link: https://lore.kernel.org/r/20220406094031.1027376-7-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add DSC for SDM845 to hw_catalog

This adds SDM845 DSC blocks into hw_catalog

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480916/
Link: https://lore.kernel.org/r/20220406094031.1027376-6-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add support for DSC in pingpong block

In SDM845, DSC can be enabled by writing to pingpong block registers, so
add support for DSC in hw_pp

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480914/
Link: https://lore.kernel.org/r/20220406094031.1027376-5-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/disp/dpu1: Add support for DSC

Display Stream Compression (DSC) is one of the hw blocks in dpu, so add
support by adding hw blocks for DSC

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480912/
Link: https://lore.kernel.org/r/20220406094031.1027376-4-vkoul@kernel.org
[DB: applied typo noticed by Robert Foss]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dsi: Pass DSC params to drm_panel

When DSC is enabled, we need to get the DSC parameters from the panel
driver, so add a dsc parameter in panel to fetch and pass DSC
configuration for DSI panels to DPU encoder, which will enable and
then configure DSC hardware blocks accordingly.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480910/
Link: https://lore.kernel.org/r/20220406094031.1027376-3-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dsi: add support for dsc data

Display Stream Compression (DSC) parameters need to be calculated. Add
helpers and struct msm_display_dsc_config in msm_drv for this
msm_display_dsc_config uses drm_dsc_config for DSC parameters.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/480908/
Link: https://lore.kernel.org/r/20220406094031.1027376-2-vkoul@kernel.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dp: enable widebus feature for display port

Widebus feature will transmit two pixel data per pixel clock to interface.
This feature now is required to be enabled to easy migrant to higher
resolution applications in future. However since some legacy chipsets
does not support this feature, this feature is enabled by setting
wide_bus_en flag to true within msm_dp_desc struct.

changes in v2:
-- remove compression related code from timing
-- remove op_info from  struct msm_drm_private
-- remove unnecessary wide_bus_en variables
-- pass wide_bus_en into timing configuration by struct msm_dp

Changes in v3:
-- split patch into 3 patches
-- enable widebus feature base on chip hardware revision

Changes in v5:
-- DP_INTF_CONFIG_DATABUS_WIDEN

Changes in v6:
-- static inline bool msm_dp_wide_bus_enable() in msm_drv.h

Changes in v7:
-- add Tested-by

Changes in v9:
-- add wide_bus_en to msm_dp_desc

Changes in v10:
-- add wide_bus_en boolean to dp_catalog struc to avoid passing it as parameter

Changes in v11:
-- add const to dp_catalog_hw_revision()
-- add const to msm_dp_wide_bus_available()

Changes in v12:
-- dp_catalog_hw_revision(const struct dp_catalog *dp_catalog)
-- msm_dp_wide_bus_available(const struct msm_dp *dp_display)

Signed-off-by: Kuogee Hsieh <quic_khsieh@quicinc.com>
Reported-by: kernel test robot <lkp@intel.com>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/476283/
Link: https://lore.kernel.org/r/1645824192-29670-5-git-send-email-quic_khsieh@quicinc.com
[DB: fixed the compilation]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: revise timing engine programming to support widebus feature

Widebus feature will transmit two pixel data per pixel clock to interface.
Timing engine provides driving force for this purpose. This patch base
on HPG (Hardware Programming Guide) to revise timing engine register
setting to accommodate both widebus and non widebus application. Also
horizontal width parameters need to be reduced by half since two pixel
data are clocked out per pixel clock when widebus feature enabled.

Widebus can be enabled individually at DP. However at DSI, widebus have
to be enabled along with DSC to achieve pixel clock rate be scaled down
with same ratio as compression ratio when 10 bits per source component.
Therefore this patch add no supports of DSI related widebus and compression.

Changes in v2:
-- remove compression related code from timing
-- remove op_info from  struct msm_drm_private
-- remove unnecessary wide_bus_en variables
-- pass wide_bus_en into timing configuration by struct msm_dp

Changes in v3:
-- split patch into 3 patches

Changes in v4:
-- rework timing engine to not interfere with dsi/hdmi
-- cover both widebus and compression

Changes in v5:
-- remove supports of DSI widebus and compression

Changes in v7:
-- split this patch into 3 patches
-- add Tested-by

Changes in v8:
-- move new registers writes under DATA_HCTL_EN features check.

Changes in v10:
-- add const inside dpu_encoder_is_widebus_enabled()
-- drop useless parenthesis please

Signed-off-by: Kuogee Hsieh <quic_khsieh@quicinc.com>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/476281/
Link: https://lore.kernel.org/r/1645824192-29670-4-git-send-email-quic_khsieh@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: replace BIT(x) with correspond marco define string

To improve code readability, this patch replace BIT(x) with
correspond register bit define string

Signed-off-by: Kuogee Hsieh <quic_khsieh@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/476280/
Link: https://lore.kernel.org/r/1645824192-29670-3-git-send-email-quic_khsieh@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: adjust display_v_end for eDP and DP

The “DP timing” requires the active region to be defined in the
bottom-right corner of the frame dimensions which is different
with DSI. Therefore both display_h_end and display_v_end need
to be adjusted accordingly. However current implementation has
only display_h_end adjusted.

Signed-off-by: Kuogee Hsieh <quic_khsieh@quicinc.com>
Fixes: fc3a69ec68d3 ("drm/msm/dpu: intf timing path for displayport")
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/476277/
Link: https://lore.kernel.org/r/1645824192-29670-2-git-send-email-quic_khsieh@quicinc.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: remove unused plane_property field from msm_drm_private

Remove unused field plane_property from struct msm_drm_private. Also
drop the enum msm_mdp_plane_property which also becomes unused.

Fixes: 7d36db0be3b9 ("drm/msm/mdp5: switch to standard zpos property")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/480939/
Link: https://lore.kernel.org/r/20220406100556.479706-1-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: don't store created planes, connectors and encoders

There is no point now in storing arrays of creates planes, connectors
and encoders. Remove them from struct msm_drm_private.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Patchwork: https://patchwork.freedesktop.org/patch/480945/
Link: https://lore.kernel.org/r/20220406101247.483649-4-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: loop over encoders using drm_for_each_encoder()

Rather than manually looping over encoders array, use standard
drm_for_each_encoder() macro.

Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/480942/
Link: https://lore.kernel.org/r/20220406101247.483649-3-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm/dpu: remove manual destruction of DRM objects

Remove manual removal of DRM modesetting objects, it is done anyway by
the drm_mode_config_cleanup() called from msm_drm_uninit(). Other
MSM display drivers (MDP4, MDP5) do not manually destroy objects and
trust generic code to do it's work.

Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/480943/
Link: https://lore.kernel.org/r/20220406101247.483649-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: properly add and remove internal bridges

Add calls to drm_bridge_add()/drm_bridge_remove() DRM bridges created by
the driver. This fixes the following warning.

WARNING: CPU: 0 PID: 1 at kernel/locking/mutex.c:579 __mutex_lock+0x840/0x9f4
DEBUG_LOCKS_WARN_ON(lock->magic != lock)
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.18.0-rc1-00002-g3054695a0d27-dirty #55
Hardware name: Generic DT based system
 unwind_backtrace from show_stack+0x10/0x14
 show_stack from dump_stack_lvl+0x58/0x70
 dump_stack_lvl from __warn+0xc8/0x1e8
 __warn from warn_slowpath_fmt+0x78/0xa8
 warn_slowpath_fmt from __mutex_lock+0x840/0x9f4
 __mutex_lock from mutex_lock_nested+0x1c/0x24
 mutex_lock_nested from drm_bridge_hpd_enable+0x2c/0x84
 drm_bridge_hpd_enable from msm_hdmi_modeset_init+0xc0/0x21c
 msm_hdmi_modeset_init from mdp4_kms_init+0x53c/0x90c
 mdp4_kms_init from msm_drm_bind+0x514/0x698
 msm_drm_bind from try_to_bring_up_aggregate_device+0x160/0x1bc
 try_to_bring_up_aggregate_device from component_master_add_with_match+0xc4/0xf8
 component_master_add_with_match from msm_pdev_probe+0x274/0x350
 msm_pdev_probe from platform_probe+0x5c/0xbc
 platform_probe from really_probe.part.0+0x9c/0x290
 really_probe.part.0 from __driver_probe_device+0xa8/0x13c
 __driver_probe_device from driver_probe_device+0x34/0x10c
 driver_probe_device from __driver_attach+0xbc/0x178
 __driver_attach from bus_for_each_dev+0x74/0xc0
 bus_for_each_dev from bus_add_driver+0x160/0x1e4
 bus_add_driver from driver_register+0x88/0x118
 driver_register from do_one_initcall+0x6c/0x334
 do_one_initcall from kernel_init_freeable+0x1bc/0x220
 kernel_init_freeable from kernel_init+0x18/0x12c
 kernel_init from ret_from_fork+0x14/0x2c

Fixes: 3d3f8b1f8b62 ("drm/bridge: make bridge registration independent of drm flow")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/481778/
Link: https://lore.kernel.org/r/20220411234953.2425280-1-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: make mdp5/dpu devices master components

The msm_mdss serves several roles at this moment. It provides IRQ domain
used by MDP5 and DPU drivers but it also serves as a component master
for both those usecases. MDP4 (which does not have separate MDSS device)
is the component master on it's own.
Remove this assymmetry and make both MDP5 and DPU component masters too.
This removes a need to care about drm/components from msm_mdss driver,
removes an mdss pointer from struct msm_drm_private and simplifies the
interface between mdp5/dpu and msm_drv.

Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482512/
Link: https://lore.kernel.org/r/20220419155346.1272627-7-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: allow compile time selection of driver components

MSM DRM driver already allows one to compile out the DP or DSI support.
Add support for disabling other features like MDP4/MDP5/DPU drivers or
direct HDMI output support.

Suggested-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482508/
Link: https://lore.kernel.org/r/20220419155346.1272627-6-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: stop using device's match data pointer

Let's make the match's data pointer a (sub-)driver's private data. The
only user currently is the msm_drm_init() function, using this data to
select kms_init callback. Pass this callback through the driver's
private data instead.

Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482510/
Link: https://lore.kernel.org/r/20220419155346.1272627-5-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: split the main platform driver

Currently the msm platform driver is a multiplex handling several cases:
- headless GPU-only driver,
- MDP4 with flat device nodes,
- MDP5/DPU MDSS with all the nodes being children of MDSS node.

This results in not-so-perfect code, checking the hardware version
(MDP4/MDP5/DPU) in several places, checking for mdss even when it can
not exist, etc. Split the code into three handling subdrivers (mdp4,
mdss and headless msm).

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482507/
Link: https://lore.kernel.org/r/20220419155346.1272627-4-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: remove extra indirection for msm_mdss

Since now there is just one mdss subdriver, drop all the indirection,
make msm_mdss struct completely opaque (and defined inside msm_mdss.c)
and call mdss functions directly.

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482505/
Link: https://lore.kernel.org/r/20220419155346.1272627-3-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: unify MDSS drivers

MDP5 and DPU1 both provide the driver handling the MDSS region, which
handles the irq domain and (incase of DPU1) adds some init for the UBWC
controller. Unify those two pieces of code into a common driver.

Reviewed-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Patchwork: https://patchwork.freedesktop.org/patch/482506/
Link: https://lore.kernel.org/r/20220419155346.1272627-2-dmitry.baryshkov@linaro.org
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>

drm/msm: return the average load over the polling period

simple_ondemand interacts poorly with clamp_to_idle.  It only looks at
the load since the last get_dev_status call, while it should really look
at the load over polling_ms.  When clamp_to_idle true, it almost always
picks the lowest frequency on active because the gpu is idle between
msm_devfreq_idle/msm_devfreq_active.

This logic could potentially be moved into devfreq core.

Fixes: 7c0ffcd40b16 ("drm/msm/gpu: Respect PM QoS constraints")
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Cc: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220416003314.59211-3-olvaffe@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: simplify gpu_busy callback

Move tracking and busy time calculation to msm_devfreq_get_dev_status.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Cc: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220416003314.59211-2-olvaffe@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: remove explicit devfreq status reset

It is redundant since commit 7c0ffcd40b16 ("drm/msm/gpu: Respect PM QoS
constraints") because dev_pm_qos_update_request triggers get_dev_status.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Cc: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220416003314.59211-1-olvaffe@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Add a way for userspace to allocate GPU iova

The motivation at this point is mainly native userspace mesa driver in a
VM guest.  The one remaining synchronous "hotpath" is buffer allocation,
because guest needs to wait to know the bo's iova before it can start
emitting cmdstream/state that references the new bo.  By allocating the
iova in the guest userspace, we no longer need to wait for a response
from the host, but can just rely on the allocation request being
processed before the cmdstream submission.  Allocation failures (OoM,
etc) would just be treated as context-lost (ie. GL_GUILTY_CONTEXT_RESET)
or subsequent allocations (or readpix, etc) can raise GL_OUT_OF_MEMORY.

v2: Fix inuse check
v3: Change mismatched iova case to -EBUSY

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Link: https://lore.kernel.org/r/20220411215849.297838-11-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Add fenced vma unpin

With userspace allocated iova (next patch), we can have a race condition
where userspace observes the fence completion and deletes the vma before
retire_submit() gets around to unpinning the vma.  To handle this, add a
fenced unpin which drops the refcount but tracks the fence, and update
msm_gem_vma_inuse() to check any previously unsignaled fences.

v2: Fix inuse underflow (duplicate unpin)
v3: Fix msm_job_run() vs submit_cleanup() race condition

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220411215849.297838-10-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Split vma lookup and pin

This way we only lookup vma once per object per submit, for both the
submit and retire path.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220411215849.297838-9-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Rework vma lookup and pin

Combines duplicate vma lookup in the get_and_pin path.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Link: https://lore.kernel.org/r/20220411215849.297838-8-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Drop msm_gem_iova()

There was only a single user, which could just as easily stash the iova
when pinning.

v2: fix prepare->prepare->cleanup->cleanup sequences

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20220411215849.297838-7-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Drop PAGE_SHIFT for address space mm

Get rid of all the unnecessary conversion between address/size and page
offsets.  It just confuses things.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20220411215849.297838-6-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Split out inuse helper

Prep for a following patch, where it gets a bit more complicated.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20220411215849.297838-5-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Convert some missed GEM_WARN_ON()s

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20220411215849.297838-4-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gpu: Drop duplicate fence counter

The ring seqno counter duplicates the fence-context last_fence counter.
They end up getting incremented in lock-step, on the same scheduler
thread, but the split just makes things less obvious.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220411215849.297838-3-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm/gem: Move prototypes

These belong more cleanly in the gem header.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20220411215849.297838-2-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Add a way to override processes comm/cmdline

In the cause of using the GPU via virtgpu, the host side process is
really a sort of proxy, and not terribly interesting from the PoV of
crash/fault logging.  Add a way to override these per process so that
we can see the guest process's name.

v2: Handle kmalloc failure, add comment to explain kstrdup returns
    NULL if passed NULL [Dan Carpenter]

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220317165144.222101-4-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Split out helper to get comm/cmdline

Deduplicate this from fault_worker and recover_worker.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220317165144.222101-3-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Add support for pointer params

The 64b value field is already suffient to hold a pointer instead of
immediate, but we also need a length field.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220317165144.222101-2-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/msm: Remove unused field in submit

Noticed this was unused and never set.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20220225202614.225197-2-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

drm/fourcc: Add QCOM tiled modifiers

These are mainly used internally in mesa, although I believe the display
should be able to scan out the TILED3 format.  Currently we define this
modifier internally in mesa for use with modifier based allocation.  But
we can get rid of that hack if we define the modfiers properly.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Link: https://lore.kernel.org/r/20210904170603.1739137-1-robdclark@gmail.com
Signed-off-by: Rob Clark <robdclark@chromium.org>

Linux 5.18-rc2

Merge tag 'tty-5.18-rc2' of git://git./linux/kernel/git/gregkh/tty

Pull serial driver fix from Greg KH:
 "This is a single serial driver fix for a build issue that showed up
  due to changes that came in through the tty tree in 5.18-rc1 that were
  missed previously. It resolves a build error with the mpc52xx_uart
  driver.

  It has been in linux-next this week with no reported problems"

* tag 'tty-5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  tty: serial: mpc52xx_uart: make rx/tx hooks return unsigned, part II.

Merge tag 'staging-5.18-rc2' of git://git./linux/kernel/git/gregkh/staging

Pull staging driver fix from Greg KH:
 "Here is a single staging driver fix for 5.18-rc2 that resolves an
  endian issue for the r8188eu driver. It has been in linux-next all
  this week with no reported problems"

* tag 'staging-5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
  staging: r8188eu: Fix PPPoE tag insertion on little endian systems

Merge tag 'driver-core-5.18-rc2' of git://git./linux/kernel/git/gregkh/driver-core

Pull driver core updates from Greg KH:
 "Here are two small driver core changes for 5.18-rc2.

  They are the final bits in the removal of the default_attrs field in
  struct kobj_type. I had to wait until after 5.18-rc1 for all of the
  changes to do this came in through different development trees, and
  then one new user snuck in. So this series has two changes:

   - removal of the default_attrs field in the powerpc/pseries/vas code.

     The change has been acked by the PPC maintainers to come through
     this tree

   - removal of default_attrs from struct kobj_type now that all
     in-kernel users are removed.

     This cleans up the kobject code a little bit and removes some
     duplicated functionality that confused people (now there is only
     one way to do default groups)

  Both of these have been in linux-next for all of this week with no
  reported problems"

* tag 'driver-core-5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
  kobject: kobj_type: remove default_attrs
  powerpc/pseries/vas: use default_groups in kobj_type

Merge tag 'char-misc-5.18-rc2' of git://git./linux/kernel/git/gregkh/char-misc

Pull char/misc driver fix from Greg KH:
 "A single driver fix. It resolves the build warning issue on 32bit
  systems in the habannalabs driver that came in during the 5.18-rc1
  merge cycle.

  It has been in linux-next for all this week with no reported problems"

* tag 'char-misc-5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  habanalabs: Fix test build failures

Merge tag 'powerpc-5.18-2' of git://git./linux/kernel/git/powerpc/linux

Pull powerpc fixes from Michael Ellerman:

 - Fix KVM "lost kick" race, where an attempt to pull a vcpu out of the
   guest could be lost (or delayed until the next guest exit).

 - Disable SCV (system call vectored) when PR KVM guests could be run.

 - Fix KVM PR guests using SCV, by disallowing AIL != 0 for KVM PR
   guests.

 - Add a new KVM CAP to indicate if AIL == 3 is supported.

 - Fix a regression when hotplugging a CPU to a memoryless/cpuless node.

 - Make virt_addr_valid() stricter for 64-bit Book3E & 32-bit, which
   fixes crashes seen due to hardened usercopy.

 - Revert a change to max_mapnr which broke HIGHMEM.

Thanks to Christophe Leroy, Fabiano Rosas, Kefeng Wang, Nicholas Piggin,
and Srikar Dronamraju.

* tag 'powerpc-5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  Revert "powerpc: Set max_mapnr correctly"
  powerpc: Fix virt_addr_valid() for 64-bit Book3E & 32-bit
  KVM: PPC: Move kvmhv_on_pseries() into kvm_ppc.h
  powerpc/numa: Handle partially initialized numa nodes
  powerpc/64: Fix build failure with allyesconfig in book3s_64_entry.S
  KVM: PPC: Use KVM_CAP_PPC_AIL_MODE_3
  KVM: PPC: Book3S PR: Disallow AIL != 0
  KVM: PPC: Book3S PR: Disable SCV when AIL could be disabled
  KVM: PPC: Book3S HV P9: Fix "lost kick" race

Merge tag 'irq-urgent-2022-04-10' of git://git./linux/kernel/git/tip/tip

Pull irq fixes from Thomas Gleixner:
 "A set of interrupt chip driver fixes:

   - A fix for a long standing bug in the ARM GICv3 redistributor
     polling which uses the wrong bit number to test.

   - Prevent translation of bogus ACPI table entries which map device
     interrupts into the IPI space on ARM GICs.

   - Don't write into the pending register of ARM GICV4 before the scan
     in hardware has completed.

   - A set of build and correctness fixes for the Qualcomm MPM driver"

* tag 'irq-urgent-2022-04-10' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  irqchip/gic, gic-v3: Prevent GSI to SGI translations
  irqchip/gic-v3: Fix GICR_CTLR.RWP polling
  irqchip/gic-v4: Wait for GICR_VPENDBASER.Dirty to clear before descheduling
  irqchip/irq-qcom-mpm: fix return value check in qcom_mpm_init()
  irq/qcom-mpm: Fix build error without MAILBOX

Merge tag 'x86_urgent_for_v5.18_rc2' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Borislav Petkov:

 - Fix the MSI message data struct definition

 - Use local labels in the exception table macros to avoid symbol
   conflicts with clang LTO builds

 - A couple of fixes to objtool checking of the relatively newly added
   SLS and IBT code

 - Rename a local var in the WARN* macro machinery to prevent shadowing

* tag 'x86_urgent_for_v5.18_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/msi: Fix msi message data shadow struct
  x86/extable: Prefer local labels in .set directives
  x86,bpf: Avoid IBT objtool warning
  objtool: Fix SLS validation for kcov tail-call replacement
  objtool: Fix IBT tail-call detection
  x86/bug: Prevent shadowing in __WARN_FLAGS
  x86/mm/tlb: Revert retpoline avoidance approach

Merge tag 'perf_urgent_for_v5.18_rc2' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Borislav Petkov:

 - A couple of fixes to cgroup-related handling of perf events

 - A couple of fixes to event encoding on Sapphire Rapids

 - Pass event caps of inherited events so that perf doesn't fail wrongly
   at fork()

 - Add support for a new Raptor Lake CPU

* tag 'perf_urgent_for_v5.18_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/core: Always set cpuctx cgrp when enable cgroup event
  perf/core: Fix perf_cgroup_switch()
  perf/core: Use perf_cgroup_info->active to check if cgroup is active
  perf/core: Don't pass task around when ctx sched in
  perf/x86/intel: Update the FRONTEND MSR mask on Sapphire Rapids
  perf/x86/intel: Don't extend the pseudo-encoding to GP counters
  perf/core: Inherit event_caps
  perf/x86/uncore: Add Raptor Lake uncore support
  perf/x86/msr: Add Raptor Lake CPU support
  perf/x86/cstate: Add Raptor Lake support
  perf/x86: Add Intel Raptor Lake support

Merge tag 'locking_urgent_for_v5.18_rc2' of git://git./linux/kernel/git/tip/tip

Pull locking fixes from Borislav Petkov:

 - Allow the compiler to optimize away unused percpu accesses and change
   the local_lock_* macros back to inline functions

 - A couple of fixes to static call insn patching

* tag 'locking_urgent_for_v5.18_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Revert "mm/page_alloc: mark pagesets as __maybe_unused"
  Revert "locking/local_lock: Make the empty local_lock_*() function a macro."
  x86/percpu: Remove volatile from arch_raw_cpu_ptr().
  static_call: Remove __DEFINE_STATIC_CALL macro
  static_call: Properly initialise DEFINE_STATIC_CALL_RET0()
  static_call: Don't make __static_call_return0 static
  x86,static_call: Fix __static_call_return0 for i386

Merge tag 'sched_urgent_for_v5.18_rc2' of git://git./linux/kernel/git/tip/tip

Pull scheduler fixes from Borislav Petkov:

 - Use the correct static key checking primitive on the IRQ exit path

 - Two fixes for the new forceidle balancer

* tag 'sched_urgent_for_v5.18_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  entry: Fix compile error in dynamic_irqentry_exit_cond_resched()
  sched: Teach the forced-newidle balancer about CPU affinity limitation.
  sched/core: Fix forceidle balancing

Merge tag 'perf-tools-fixes-for-v5.18-2022-04-09' of git://git./linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Fix the clang command line option probing and remove some options to
   filter out, fixing the build with the latest clang versions

 - Fix 'perf bench' futex and epoll benchmarks to deal with machines
   with more than 1K CPUs

 - Fix 'perf test tsc' error message when not supported

 - Remap perf ring buffer if there is no space for event, fixing perf
   usage in 32-bit ChromeOS

 - Drop objdump stderr to avoid getting stuck waiting for stdout output
   in 'perf annotate'

 - Fix up garbled output by now showing unwind error messages when
   augmenting frame in best effort mode

 - Fix perf's libperf_print callback, use the va_args eprintf() variant

 - Sync vhost and arm64 cputype headers with the kernel sources

 - Fix 'perf report --mem-mode' with ARM SPE

 - Add missing external commands ('iiostat', etc) to 'perf --list-cmds'

* tag 'perf-tools-fixes-for-v5.18-2022-04-09' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  perf annotate: Drop objdump stderr to avoid getting stuck waiting for stdout output
  perf tools: Add external commands to list-cmds
  perf docs: Add perf-iostat link to manpages
  perf session: Remap buf if there is no space for event
  perf bench: Fix epoll bench to correct usage of affinity for machines with #CPUs > 1K
  perf bench: Fix futex bench to correct usage of affinity for machines with #CPUs > 1K
  perf tools: Fix perf's libperf_print callback
  perf: arm-spe: Fix perf report --mem-mode
  perf unwind: Don't show unwind error messages when augmenting frame pointer stack
  tools headers arm64: Sync arm64's cputype.h with the kernel sources
  perf test tsc: Fix error message when not supported
  perf build: Don't use -ffat-lto-objects in the python feature test when building with clang-13
  perf python: Fix probing for some clang command line options
  tools build: Filter out options and warnings not supported by clang
  tools build: Use $(shell ) instead of `` to get embedded libperl's ccopts
  tools include UAPI: Sync linux/vhost.h with the kernel sources

Merge tag 'cxl+nvdimm-for-5.18-rc2' of git://git./linux/kernel/git/nvdimm/nvdimm

Pull cxl and nvdimm fixes from Dan Williams:

 - Fix a compile error in the nvdimm unit tests

 - Fix a shadowed variable warning in the CXL PCI driver

* tag 'cxl+nvdimm-for-5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
  cxl/pci: Drop shadowed variable
  tools/testing/nvdimm: Fix security_init() symbol collision

Merge tag 'gpio-fixes-for-v5.18-rc2' of git://git./linux/kernel/git/brgl/linux

Pull gpio fix from Bartosz Golaszewski:

 - fix a race condition with consumers accessing the fields of GPIO IRQ
   chips before they're fully initialized

* tag 'gpio-fixes-for-v5.18-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:
  gpio: Restrict usage of GPIO chip irq members before initialization

Merge tag 'irqchip-fixes-5.18-1' of git://git./linux/kernel/git/maz/arm-platforms into irq/urgent

Pull irqchip fixes from Marc Zyngier:

 - Fix GICv3 polling for RWP in redistributors

 - Reject ACPI attempts to use SGIs on GIC/GICv3

 - Fix unpredictible behaviour when making a VPE non-resident
   with GICv4

 - A couple of fixes for the newly merged qcom-mpm driver

Link: https://lore.kernel.org/lkml/20220409094229.267649-1-maz@kernel.org

perf annotate: Drop objdump stderr to avoid getting stuck waiting for stdout output

If objdump writes to stderr it can block waiting for it to be read. As
perf doesn't read stderr then progress stops with perf waiting for
stdout output.

Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexandre Truong <alexandre.truong@arm.com>
Cc: Dave Marchevsky <davemarchevsky@fb.com>
Cc: Denis Nikitin <denik@chromium.org>
Cc: German Gomez <german.gomez@arm.com>
Cc: James Clark <james.clark@arm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Garry <john.garry@huawei.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Lexi Shao <shaolexi@huawei.com>
Cc: Li Huafei <lihuafei1@huawei.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Martin Liška <mliska@suse.cz>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@amd.com>
Cc: Remi Bernon <rbernon@codeweavers.com>
Cc: Riccardo Mancini <rickyman7@gmail.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Will Deacon <will@kernel.org>
Cc: William Cohen <wcohen@redhat.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lore.kernel.org/lkml/20220407230503.1265036-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Add external commands to list-cmds

The `perf --list-cmds` output prints only internal commands, although
there is no reason for that from users' perspective.

Adding the external commands to commands array with NULL function
pointer allows printing all perf commands while not changing the logic
of command handler selection.

Signed-off-by: Michael Petlan <mpetlan@redhat.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20220404221541.30312-2-mpetlan@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf docs: Add perf-iostat link to manpages

Signed-off-by: Michael Petlan <mpetlan@redhat.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20220404221541.30312-1-mpetlan@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf session: Remap buf if there is no space for event

If a perf event doesn't fit into remaining buffer space return NULL to
remap buf and fetch the event again.

Keep the logic to error out on inadequate input from fuzzing.

This fixes perf failing on ChromeOS (with 32b userspace):

  $ perf report -v -i perf.data
  ...
  prefetch_event: head=0x1fffff8 event->header_size=0x30, mmap_size=0x2000000: fuzzed or compressed perf.data?
  Error:
  failed to process sample

Fixes: 57fc032ad643ffd0 ("perf session: Avoid infinite loop when seeing invalid header.size")
Reviewed-by: James Clark <james.clark@arm.com>
Signed-off-by: Denis Nikitin <denik@chromium.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20220330031130.2152327-1-denik@chromium.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'scsi-fixes' of git://git./linux/kernel/git/jejb/scsi

Pull SCSI fixes from James Bottomley:

 - add support for new devices (ufs, mvsas)

 - a major set of fixes in lpfc

 - get rid of a driver specific ioctl in pcmraid

 - a major rework of aha152x to get rid of the scsi_pointer.

 - minor fixes and obvious changes including several spelling updates.

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (36 commits)
  scsi: megaraid_sas: Target with invalid LUN ID is deleted during scan
  scsi: ufs: ufshpb: Fix a NULL check on list iterator
  scsi: sd: Clean up gendisk if device_add_disk() failed
  scsi: message: fusion: Remove redundant variable dmp
  scsi: mvsas: Add PCI ID of RocketRaid 2640
  scsi: sd: sd_read_cpr() requires VPD pages
  scsi: mpt3sas: Fail reset operation if config request timed out
  scsi: sym53c500_cs: Stop using struct scsi_pointer
  scsi: ufs: ufs-pci: Add support for Intel MTL
  scsi: mpt3sas: Fix mpt3sas_check_same_4gb_region() kdoc comment
  scsi: scsi_debug: Fix sdebug_blk_mq_poll() in_use_bm bitmap use
  scsi: bnx2i: Fix spelling mistake "mis-match" -> "mismatch"
  scsi: bnx2fc: Fix spelling mistake "mis-match" -> "mismatch"
  scsi: zorro7xx: Fix a resource leak in zorro7xx_remove_one()
  scsi: aic7xxx: Use standard PCI subsystem, subdevice defines
  scsi: ufs: qcom: Drop custom Android boot parameters
  scsi: core: sysfs: Remove comments that conflict with the actual logic
  scsi: hisi_sas: Remove stray fallthrough annotation
  scsi: virtio-scsi: Eliminate anonymous module_init & module_exit
  scsi: isci: Fix spelling mistake "doesnt" -> "doesn't"
  ...

perf bench: Fix epoll bench to correct usage of affinity for machines with #CPUs > 1K

The 'perf bench epoll' testcase fails on systems with more than 1K CPUs.

Testcase: perf bench epoll all

Result snippet:
<<>>
Run summary [PID 106497]: 1399 threads monitoring on 64 file-descriptors for 8 secs.

perf: pthread_create: No such file or directory
<<>>

In epoll benchmarks (ctl, wait) pthread_create is invoked in do_threads
from respective bench_epoll_*  function. Though the logs shows direct
failure from pthread_create, the actual failure is from
"sched_setaffinity" returning EINVAL (invalid argument).

This happens because the default mask size in glibc is 1024. To overcome
this 1024 CPUs mask size limitation of cpu_set_t, change the mask size
using the CPU_*_S macros.

Patch addresses this by fixing all the epoll benchmarks to use CPU_ALLOC
to allocate cpumask, CPU_ALLOC_SIZE for size, and CPU_SET_S to set the
mask.

Reported-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Signed-off-by: Athira Jajeev <atrajeev@linux.vnet.ibm.com>
Tested-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Nageswara R Sastry <rnsastry@linux.ibm.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Link: https://lore.kernel.org/r/20220406175113.87881-3-atrajeev@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf bench: Fix futex bench to correct usage of affinity for machines with #CPUs > 1K

The 'perf bench futex' testcase fails on systems with more than 1K CPUs.

Testcase: perf bench futex all

Failure snippet:
<<>>Running futex/hash benchmark...

perf: pthread_create: No such file or directory
<<>>

All the futex benchmarks (ie hash, lock-api, requeue, wake,
wake-parallel), pthread_create is invoked in respective bench_futex_*
function. Though the logs shows direct failure from pthread_create,
strace logs showed that actual failure is from  "sched_setaffinity"
returning EINVAL (invalid argument).

This happens because the default mask size in glibc is 1024. To overcome
this 1024 CPUs mask size limitation of cpu_set_t, change the mask size
using the CPU_*_S macros.

Patch addresses this by fixing all the futex benchmarks to use CPU_ALLOC
to allocate cpumask, CPU_ALLOC_SIZE for size, and CPU_SET_S to set the
mask.

Reported-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Athira Jajeev <atrajeev@linux.vnet.ibm.com>
Tested-by: Disha Goel <disgoel@linux.vnet.ibm.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Nageswara R Sastry <rnsastry@linux.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Link: https://lore.kernel.org/r/20220406175113.87881-2-atrajeev@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Fix perf's libperf_print callback

eprintf() does not expect va_list as the type of the 4th parameter.

Use veprintf() because it does.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Fixes: 428dab813a56ce94 ("libperf: Merge libperf_set_print() into libperf_init()")
Cc: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20220408132625.2451452-1-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf: arm-spe: Fix perf report --mem-mode

Since commit bb30acae4c4dacfa ("perf report: Bail out --mem-mode if mem
info is not available") "perf mem report" and "perf report --mem-mode"
don't allow opening the file unless one of the events has
PERF_SAMPLE_DATA_SRC set.

SPE doesn't have this set even though synthetic memory data is generated
after it is decoded. Fix this issue by setting DATA_SRC on SPE events.
This has no effect on the data collected because the SPE driver doesn't
do anything with that flag and doesn't generate samples.

Fixes: bb30acae4c4dacfa ("perf report: Bail out --mem-mode if mem info is not available")
Signed-off-by: James Clark <james.clark@arm.com>
Tested-by: Leo Yan <leo.yan@linaro.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: German Gomez <german.gomez@arm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Garry <john.garry@huawei.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20220408144056.1955535-1-james.clark@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf unwind: Don't show unwind error messages when augmenting frame pointer stack

Commit Fixes: b9f6fbb3b2c29736 ("perf arm64: Inject missing frames when
using 'perf record --call-graph=fp'") intended to add a 'best effort'
DWARF unwind that improved the frame pointer stack in most scenarios.

It's expected that the unwind will fail sometimes, but this shouldn't be
reported as an error. It only works when the return address can be
determined from the contents of the link register alone.

Fix the error shown when the unwinder requires extra registers by adding
a new flag that suppresses error messages. This flag is not set in the
normal --call-graph=dwarf unwind mode so that behavior is not changed.

Fixes: b9f6fbb3b2c29736 ("perf arm64: Inject missing frames when using 'perf record --call-graph=fp'")
Reported-by: John Garry <john.garry@huawei.com>
Signed-off-by: James Clark <james.clark@arm.com>
Tested-by: John Garry <john.garry@huawei.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexandre Truong <alexandre.truong@arm.com>
Cc: German Gomez <german.gomez@arm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20220406145651.1392529-1-james.clark@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools headers arm64: Sync arm64's cputype.h with the kernel sources

To get the changes in:

  83bea32ac7ed37bb ("arm64: Add part number for Arm Cortex-A78AE")

That addresses this perf build warning:

  Warning: Kernel ABI header at 'tools/arch/arm64/include/asm/cputype.h' differs from latest version at 'arch/arm64/include/asm/cputype.h'
  diff -u tools/arch/arm64/include/asm/cputype.h arch/arm64/include/asm/cputype.h

Cc: Ali Saidi <alisaidi@amazon.com>
Cc: Andrew Kilroy <andrew.kilroy@arm.com>
Cc: Chanho Park <chanho61.park@samsung.com>
Cc: German Gomez <german.gomez@arm.com>
Cc: James Clark <james.clark@arm.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Will Deacon <will@kernel.org>
Link: http://lore.kernel.org/lkml/
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf test tsc: Fix error message when not supported

By default `perf test tsc` does not return the error message when the
child process detected kernel does not support it. Instead, the child
process prints an error message to stderr, unfortunately stderr is
redirected to /dev/null when verbose <= 0.

This patch does:

- return TEST_SKIP to the parent process instead of TEST_OK when
  perf_read_tsc_conversion() is not supported.

- Add a new subtest of testing if TSC is supported on current
  architecture by moving exist code to a separate function.
  It avoids two places in test__perf_time_to_tsc() that return
  TEST_SKIP by doing this.

- Extend the test suite definition to contain above two subtests.
  Current test_suite and test_case structs do not support printing skip
  reason when the number of subtest less than 1. To print skip reason, it
  is necessary to extend current test suite definition.

Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Chengdong Li <chengdongli@tencent.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: likexu@tencent.com
Link: https://lore.kernel.org/r/20220408084748.43707-1-chengdongli@tencent.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf build: Don't use -ffat-lto-objects in the python feature test when building with clang-13

Using -ffat-lto-objects in the python feature test when building with
clang-13 results in:

  clang-13: error: optimization flag '-ffat-lto-objects' is not supported [-Werror,-Wignored-optimization-argument]
  error: command '/usr/sbin/clang' failed with exit code 1
  cp: cannot stat '/tmp/build/perf/python_ext_build/lib/perf*.so': No such file or directory
  make[2]: *** [Makefile.perf:639: /tmp/build/perf/python/perf.so] Error 1

Noticed when building on a docker.io/library/archlinux:base container.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Fangrui Song <maskray@google.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Keeping <john@metanate.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf python: Fix probing for some clang command line options

The clang compiler complains about some options even without a source
file being available, while others require one, so use the simple
tools/build/feature/test-hello.c file.

Then check for the "is not supported" string in its output, in addition
to the "unknown argument" already being looked for.

This was noticed when building with clang-13 where -ffat-lto-objects
isn't supported and since we were looking just for "unknown argument"
and not providing a source code to clang, was mistakenly assumed as
being available and not being filtered to set of command line options
provided to clang, leading to a build failure.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Fangrui Song <maskray@google.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Keeping <john@metanate.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Sedat Dilek <sedat.dilek@gmail.com>
Link: http://lore.kernel.org/lkml/
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools build: Filter out options and warnings not supported by clang

These make the feature check fail when using clang, so remove them just
like is done in tools/perf/Makefile.config to build perf itself.

Adding -Wno-compound-token-split-by-macro to tools/perf/Makefile.config
when building with clang is also necessary to avoid these warnings
turned into errors (-Werror):

    CC      /tmp/build/perf/util/scripting-engines/trace-event-perl.o
  In file included from util/scripting-engines/trace-event-perl.c:35:
  In file included from /usr/lib64/perl5/CORE/perl.h:4085:
  In file included from /usr/lib64/perl5/CORE/hv.h:659:
  In file included from /usr/lib64/perl5/CORE/hv_func.h:34:
  In file included from /usr/lib64/perl5/CORE/sbox32_hash.h:4:
  /usr/lib64/perl5/CORE/zaphod32_hash.h:150:5: error: '(' and '{' tokens introducing statement expression appear in different macro expansion contexts [-Werror,-Wcompound-token-split-by-macro]
      ZAPHOD32_SCRAMBLE32(state[0],0x9fade23b);
      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/lib64/perl5/CORE/zaphod32_hash.h:80:38: note: expanded from macro 'ZAPHOD32_SCRAMBLE32'
  #define ZAPHOD32_SCRAMBLE32(v,prime) STMT_START {  \
                                       ^~~~~~~~~~
  /usr/lib64/perl5/CORE/perl.h:737:29: note: expanded from macro 'STMT_START'
  #   define STMT_START   (void)( /* gcc supports "({ STATEMENTS; })" */
                                ^
  /usr/lib64/perl5/CORE/zaphod32_hash.h:150:5: note: '{' token is here
      ZAPHOD32_SCRAMBLE32(state[0],0x9fade23b);
      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/lib64/perl5/CORE/zaphod32_hash.h:80:49: note: expanded from macro 'ZAPHOD32_SCRAMBLE32'
  #define ZAPHOD32_SCRAMBLE32(v,prime) STMT_START {  \
                                                  ^
  /usr/lib64/perl5/CORE/zaphod32_hash.h:150:5: error: '}' and ')' tokens terminating statement expression appear in different macro expansion contexts [-Werror,-Wcompound-token-split-by-macro]
      ZAPHOD32_SCRAMBLE32(state[0],0x9fade23b);
      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/lib64/perl5/CORE/zaphod32_hash.h:87:41: note: expanded from macro 'ZAPHOD32_SCRAMBLE32'
      v ^= (v>>23);                       \
                                          ^
  /usr/lib64/perl5/CORE/zaphod32_hash.h:150:5: note: ')' token is here
      ZAPHOD32_SCRAMBLE32(state[0],0x9fade23b);
      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  /usr/lib64/perl5/CORE/zaphod32_hash.h:88:3: note: expanded from macro 'ZAPHOD32_SCRAMBLE32'
  } STMT_END
    ^~~~~~~~
  /usr/lib64/perl5/CORE/perl.h:738:21: note: expanded from macro 'STMT_END'
  #   define STMT_END     )
                          ^

Please refer to the discussion on the Link: tag below, where Nathan
clarifies the situation:

<quote>
acme> And then get to the problems at the end of this message, which seem
acme> similar to the problem described here:
acme>
acme> From  Nathan Chancellor <>
acme> Subject [PATCH] mwifiex: Remove unnecessary braces from HostCmd_SET_SEQ_NO_BSS_INFO
acme>
acme> https://lkml.org/lkml/2020/9/1/135
acme>
acme> So perhaps in this case its better to disable that
acme> -Werror,-Wcompound-token-split-by-macro when building with clang?

Yes, I think that is probably the best solution. As far as I can tell,
at least in this file and context, the warning appears harmless, as the
"create a GNU C statement expression from two different macros" is very
much intentional, based on the presence of PERL_USE_GCC_BRACE_GROUPS.
The warning is fixed in upstream Perl by just avoiding creating GNU C
statement expressions using STMT_START and STMT_END:

  https://github.com/Perl/perl5/issues/18780
  https://github.com/Perl/perl5/pull/18984

If I am reading the source code correctly, an alternative to disabling
the warning would be specifying -DPERL_GCC_BRACE_GROUPS_FORBIDDEN but it
seems like that might end up impacting more than just this site,
according to the issue discussion above.
</quote>

Based-on-a-patch-by: Sedat Dilek <sedat.dilek@gmail.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com> # Debian/Selfmade LLVM-14 (x86-64)
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Fangrui Song <maskray@google.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Keeping <john@metanate.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Link: http://lore.kernel.org/lkml/YkxWcYzph5pC1EK8@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools build: Use $(shell ) instead of `` to get embedded libperl's ccopts

Just like its done for ldopts and for both in tools/perf/Makefile.config.

Using `` to initialize PERL_EMBED_CCOPTS somehow precludes using:

  $(filter-out SOMETHING_TO_FILTER,$(PERL_EMBED_CCOPTS))

And we need to do it to allow for building with versions of clang where
some gcc options selected by distros are not available.

Tested-by: Sedat Dilek <sedat.dilek@gmail.com> # Debian/Selfmade LLVM-14 (x86-64)
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Fangrui Song <maskray@google.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: John Keeping <john@metanate.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Link: http://lore.kernel.org/lkml/YktYX2OnLtyobRYD@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools include UAPI: Sync linux/vhost.h with the kernel sources

To get the changes in:

  b04d910af330b55e ("vdpa: support exposing the count of vqs to userspace")
  a61280ddddaa45f9 ("vdpa: support exposing the config size to userspace")

Silencing this perf build warning:

  Warning: Kernel ABI header at 'tools/include/uapi/linux/vhost.h' differs from latest version at 'include/uapi/linux/vhost.h'
  diff -u tools/include/uapi/linux/vhost.h include/uapi/linux/vhost.h

  $ diff -u tools/include/uapi/linux/vhost.h include/uapi/linux/vhost.h
  --- tools/include/uapi/linux/vhost.h 2021-07-15 16:17:01.840818309 -0300
  +++ include/uapi/linux/vhost.h 2022-04-02 18:55:05.702522387 -0300
  @@ -150,4 +150,11 @@
   /* Get the valid iova range */
   #define VHOST_VDPA_GET_IOVA_RANGE _IOR(VHOST_VIRTIO, 0x78, \
         struct vhost_vdpa_iova_range)
  +
  +/* Get the config size */
  +#define VHOST_VDPA_GET_CONFIG_SIZE _IOR(VHOST_VIRTIO, 0x79, __u32)
  +
  +/* Get the count of all virtqueues */
  +#define VHOST_VDPA_GET_VQS_COUNT _IOR(VHOST_VIRTIO, 0x80, __u32)
  +
   #endif
  $ tools/perf/trace/beauty/vhost_virtio_ioctl.sh > before
  $ cp include/uapi/linux/vhost.h tools/include/uapi/linux/vhost.h
  $ tools/perf/trace/beauty/vhost_virtio_ioctl.sh > after
  $ diff -u before after
  --- before 2022-04-04 14:52:25.036375145 -0300
  +++ after 2022-04-04 14:52:31.906549976 -0300
  @@ -38,4 +38,6 @@
    [0x73] = "VDPA_GET_CONFIG",
    [0x76] = "VDPA_GET_VRING_NUM",
    [0x78] = "VDPA_GET_IOVA_RANGE",
  + [0x79] = "VDPA_GET_CONFIG_SIZE",
  + [0x80] = "VDPA_GET_VQS_COUNT",
   };
  $

Cc: Longpeng <longpeng2@huawei.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/lkml/YksxoFcOARk%2Fldev@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'block-5.18-2022-04-08' of git://git.kernel.dk/linux-block

Pull block fixes from Jens Axboe:
 "Nothing major in here, just a few small fixes:

   - Small series of neglected drbd patches (Christoph, Lv, Xiaomeng)

   - Remove dead variable in cdrom (Enze)"

* tag 'block-5.18-2022-04-08' of git://git.kernel.dk/linux-block:
  drbd: set QUEUE_FLAG_STABLE_WRITES
  drbd: fix an invalid memory access caused by incorrect use of list iterator
  drbd: Fix five use after free bugs in get_initial_state
  cdrom: remove unused variable

Merge tag 'io_uring-5.18-2022-04-08' of git://git.kernel.dk/linux-block

Pull io_uring fixes from Jens Axboe:
 "A bit bigger than usual post merge window, largely due to a revert and
  a fix of at what point files are assigned for requests.

  The latter fixing a linked request use case where a dependent link can
  rely on what file is assigned consistently.

  Summary:

   - 32-bit compat fix for IORING_REGISTER_IOWQ_AFF (Eugene)

   - File assignment fixes (me)

   - Revert of the NAPI poll addition from this merge window. The author
     isn't available right now to engage on this, so let's revert it and
     we can retry for the 5.19 release (me, Jakub)

   - Fix a timeout removal race (me)

   - File update and SCM fixes (Pavel)"

* tag 'io_uring-5.18-2022-04-08' of git://git.kernel.dk/linux-block:
  io_uring: fix race between timeout flush and removal
  io_uring: use nospec annotation for more indexes
  io_uring: zero tag on rsrc removal
  io_uring: don't touch scm_fp_list after queueing skb
  io_uring: nospec index for tags on files update
  io_uring: implement compat handling for IORING_REGISTER_IOWQ_AFF
  Revert "io_uring: Add support for napi_busy_poll"
  io_uring: drop the old style inflight file tracking
  io_uring: defer file assignment
  io_uring: propagate issue_flags state down to file assignment
  io_uring: move read/write file prep state into actual opcode handler
  io_uring: defer splice/tee file validity check until command issue
  io_uring: don't check req->file in io_fsync_prep()

Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:
 "Several bug fixes for old bugs:

   - Welcome Leon as co-maintainer for RDMA so we are back to having two
     people

   - Some corner cases are fixed in mlx5's MR code

   - Long standing CM bug where a DREQ at the wrong time can result in a
     long timeout

   - Missing locking and refcounting in hf1"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/hfi1: Fix use-after-free bug for mm struct
  IB/rdmavt: add lock to call to rvt_error_qp to prevent a race condition
  IB/cm: Cancel mad on the DREQ event when the state is MRA_REP_RCVD
  RDMA/mlx5: Add a missing update of cache->last_add
  RDMA/mlx5: Don't remove cache MRs when a delay is needed
  MAINTAINERS: Update qib and hfi1 related drivers
  MAINTAINERS: Add Leon Romanovsky to RDMA maintainers