[igt-dev] [PATCH i-g-t 5/9] tools: Add intel_pm_rpm tool
Gupta, Anshuman
anshuman.gupta at intel.com
Thu Apr 28 06:59:46 UTC 2022
> -----Original Message-----
> From: Vivi, Rodrigo <rodrigo.vivi at intel.com>
> Sent: Friday, April 22, 2022 5:38 PM
> To: Gupta, Anshuman <anshuman.gupta at intel.com>
> Cc: igt-dev at lists.freedesktop.org; Nilawar, Badal <badal.nilawar at intel.com>
> Subject: Re: [igt-dev] [PATCH i-g-t 5/9] tools: Add intel_pm_rpm tool
>
> On Fri, Apr 22, 2022 at 06:22:28AM -0400, Rodrigo Vivi wrote:
> > On Mon, Apr 18, 2022 at 06:20:44PM +0530, Anshuman Gupta wrote:
> > > intel_pm_rpm tool is a debug tool. It can be use to setup and
> > > prepare the gfx card to go to D3Cold.
> > > It also provide the debug option to disable all display and prepare
> > > device to enter to runtime suspend.
> > >
> > > Cc: Rodrigo Vivi <rodrigo.vivi at intel.com>
> > > Signed-off-by: Anshuman Gupta <anshuman.gupta at intel.com>
> > > ---
> > > tools/intel_pm_rpm.c | 209
> +++++++++++++++++++++++++++++++++++++++++++
> > > tools/meson.build | 1 +
> > > 2 files changed, 210 insertions(+)
> > > create mode 100644 tools/intel_pm_rpm.c
> > >
> > > diff --git a/tools/intel_pm_rpm.c b/tools/intel_pm_rpm.c new file
> > > mode 100644 index 000000000..df9cfa632
> > > --- /dev/null
> > > +++ b/tools/intel_pm_rpm.c
> > > @@ -0,0 +1,209 @@
> > > +/*
> > > + * Copyright (c) 2022 Intel Corporation
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > > +obtaining a
> > > + * copy of this software and associated documentation files (the
> > > +"Software"),
> > > + * to deal in the Software without restriction, including without
> > > +limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > > +sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to
> > > +whom the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including
> > > +the next
> > > + * paragraph) shall be included in all copies or substantial
> > > +portions of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY
> KIND,
> > > +EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > > +MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO
> > > +EVENT SHALL
> > > + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
> > > +DAMAGES OR OTHER
> > > + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > +ARISING
> > > + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> OR
> > > +OTHER DEALINGS
> > > + * IN THE SOFTWARE.
> > > + *
> > > + */
> > > +
> > > +#include <errno.h>
> > > +#include <getopt.h>
> > > +#include <glib.h>
> > > +#include <stdio.h>
> > > +#include <stdlib.h>
> > > +#include <string.h>
> > > +#include <stdbool.h>
> > > +#include "igt.h"
> > > +#include "igt_device.h"
> > > +#include "igt_device_scan.h"
> > > +#include "igt_pm.h"
> > > +
> > > +typedef struct {
> > > + int drm_fd;
> > > + int debugfs_fd;
> > > + uint32_t devid;
> > > + drmModeResPtr res;
> > > + igt_display_t display;
> > > +} data_t;
> > > +
> > > +const char *help_str =
> > > + " --disable-display\t\tDisable all screen and try to go into runtime
> pm.\n"
> > > + " --setup-d3cold\t\tPrepare dgfx gfx card to enter runtime D3Cold.\n"
> > > + " --help\t\tProvide help. Provide card name with
> > > +IGT_DEVICE=drm:/dev/dri/card*."; static struct option long_options[] = {
> > > + {"disable-display", 0, 0, 'd'},
> > > + {"setup-d3cold", 0, 0, 's'},
> > > + {"help", 0, 0, 'h'},
> > > + { 0, 0, 0, 0 }
> > > +};
> > > +
> > > +const char *optstr = "dsh";
> > > +
> > > +static void usage(const char *name) {
> > > + igt_info("Usage: %s [options]\n", name);
> > > + igt_info("%s\n", help_str);
> > > +}
> > > +
> > > +static void disable_all_displays(data_t *data) {
> > > + igt_output_t *output;
> > > +
> > > + for (int i = 0; i < data->display.n_outputs; i++) {
> > > + output = &data->display.outputs[i];
> > > + igt_output_set_pipe(output, PIPE_NONE);
> > > + igt_display_commit(&data->display);
> > > + }
> > > +}
> > > +
> > > +static void setup_gfx_card_d3cold(data_t *data) {
> > > + struct pci_device *root;
> > > + int d_state;
> > > +
> > > + /* igfx does not support d3cold */
> > > + if (!IS_DGFX(data->devid))
> > > + return;
> >
> > I believe the if below will already take care of this since the
> > real_power_state won't show up. However, let's keep this check here to
> > be really clear on the expectations and to avoid navigating the tree in vain...
> >
> > > +
> > > + root = igt_device_get_pci_root_port(data->drm_fd);
> > > +
> > > + if (!igt_pm_acpi_d3cold_supported(root)) {
> > > + igt_info("D3Cold isn't supported on Root port
> %04x:%02x:%02x.%01x\n",
> > > + root->domain, root->bus, root->dev, root->func);
> > > + return;
> > > + }
> > > +
> > > + disable_all_displays(data);
> > > + igt_pm_setup_pci_card_runtime_pm(root);
> > > + sleep(1);
> > > + d_state = igt_pm_get_acpi_real_d_state(root);
> > > +
> > > + if (d_state == IGT_ACPI_D3Cold) {
> > > + igt_info("D3Cold achieved for root port
> %04x:%02x:%02x.%01x\n",
> > > + root->domain, root->bus, root->dev, root->func);
> > > + } else {
> > > + igt_pm_print_pci_card_runtime_status();
> > > + igt_info("D3Cold not achieved yet. Please monitor
> %04x:%02x:%02x.%01x real_power_state\n",
> > > + root->domain, root->bus, root->dev, root->func);
>
>
> oh, I had forgot to send comment on this one...
> I believe that with sleep(1) above you are always reaching to this msg here, no?!
>
> Should we also change the autosuspend from everyone to 1 and maybe wait 2
> here?
May be 0 should be ok to write all devices auto suspend delay.
The problem is here writing 1 will conflict with existing i915 runtime pm setup layer,
Which writes 0 to i915 auto suspend delay.
Another issue I had observed while writing auto suspend delay ,
Few pci devices doesn't support auto-suspend.
----------------------------------------------------------------------------------
ta at DUT2135-DG2MRB:~/drivers.gpu.i915.igt-gpu-tools$ sudo ./build/tools/intel_pm_rpm --s
Device /dev/dri/card1 successfully opened
DMC: fw loaded: yes
Test requirement not met in function __igt_pm_setup_pci_card_runtime_pm, file ../lib/igt_pm.c:969:
Test requirement: read(delay_fd, delay, delay_size - 1) > 0
PCI '0000:02:01.0' doesn't support auto_suspend
Last errno: 5, Input/output error
----------------------------------------------------------------------------------
Br,
Anshuman Gupta.
>
> > > + }
> > > +}
> > > +
> > > +int main(int argc, char *argv[])
> > > +{
> > > + bool disable_display = false, setup_d3cold = false;
> > > + struct igt_device_card card;
> > > + char *env_device = NULL;
> > > + int c, option_index = 0;
> > > + data_t data = {};
> > > + int ret = 0;
> > > +
> > > + if (argc <= 1) {
> > > + usage(argv[0]);
> > > + return EXIT_SUCCESS;
> > > + }
> > > +
> > > + env_device = getenv("IGT_DEVICE");
> > > + igt_devices_scan(false);
> > > +
> > > + if (env_device) {
> > > + if (!igt_device_card_match(env_device, &card)) {
> > > + igt_warn("No device found for the env_device\n");
> > > + ret = EXIT_FAILURE;
> > > + goto exit;
> > > + }
> > > + } else {
> > > + if (!igt_device_find_first_i915_discrete_card(&card)) {
> > > + igt_warn("No discrete gpu found\n");
> > > + ret = EXIT_FAILURE;
> > > + goto exit;
> > > + }
> > > + }
> > > +
> > > + while ((c = getopt_long(argc, argv, optstr,
> > > + long_options, &option_index)) != -1) {
> > > + switch (c) {
> > > + case 'd':
> > > + disable_display = true;
> > > + break;
> > > + case 's':
> > > + setup_d3cold = true;
> > > + break;
> > > + default:
> > > + case 'h':
> > > + usage(argv[0]);
> > > + ret = EXIT_SUCCESS;
> > > + goto exit;
> > > + }
> > > + }
> > > +
> > > + data.drm_fd = igt_open_card(&card);
> > > + if (data.drm_fd >= 0) {
> > > + igt_info("Device %s successfully opened\n", card.card);
> > > + } else {
> > > + igt_warn("Cannot open card %s device\n", card.card);
> > > + ret = EXIT_FAILURE;
> > > + goto exit;
> > > + }
> > > +
> > > + data.debugfs_fd = igt_debugfs_dir(data.drm_fd);
> > > + data.devid = intel_get_drm_devid(data.drm_fd);
> > > + igt_setup_runtime_pm(data.drm_fd);
> > > +
> > > + data.res = drmModeGetResources(data.drm_fd);
> > > + if (data.res) {
> > > + kmstest_set_vt_graphics_mode();
> > > + igt_display_require(&data.display, data.drm_fd);
> > > +
> > > + if (!igt_pm_dmc_loaded(data.debugfs_fd)) {
> > > + igt_warn("dmc fw is not loaded, no runtime pm\n");
> > > + ret = EXIT_FAILURE;
> > > + goto exit;
> > > + }
> > > + }
> >
> > Do we really need this? It is only for checking if DMC is there in case we have
> display right?
> > Should deserve a comment at least...
> >
> > but already can use my:
> >
> > Reviewed-by: Rodrigo Vivi <rodrigo.vivi at intel.com>
> >
> > if explained with reply here or with comment if you see it fits
> >
> > > +
> > > + if (disable_display) {
> > > + disable_all_displays(&data);
> > > + if
> (!igt_wait_for_pm_status(IGT_RUNTIME_PM_STATUS_SUSPENDED)) {
> > > + __igt_debugfs_dump(data.drm_fd,
> "i915_runtime_pm_status", IGT_LOG_INFO);
> > > + ret = EXIT_FAILURE;
> > > + goto exit;
> > > + }
> > > +
> > > + igt_info("Device runtime suspended, Useful for debugging.\n"
> > > + "Hit CTRL-C to exit\n");
> > > + /* Don't return useful for debugging */
> > > + while (1)
> > > + sleep(600);
> > > + }
> > > +
> > > + if (setup_d3cold)
> > > + setup_gfx_card_d3cold(&data);
> > > +
> > > +exit:
> > > + igt_restore_runtime_pm();
> > > +
> > > + if (data.res)
> > > + igt_display_fini(&data.display);
> > > +
> > > + close(data.debugfs_fd);
> > > + close(data.drm_fd);
> > > + igt_devices_free();
> > > +
> > > + return ret;
> > > +}
> > > diff --git a/tools/meson.build b/tools/meson.build index
> > > 771d0b9e3..24d0ea714 100644
> > > --- a/tools/meson.build
> > > +++ b/tools/meson.build
> > > @@ -28,6 +28,7 @@ tools_progs = [
> > > 'intel_lid',
> > > 'intel_opregion_decode',
> > > 'intel_panel_fitter',
> > > + 'intel_pm_rpm',
> > > 'intel_reg_checker',
> > > 'intel_residency',
> > > 'intel_stepping',
> > > --
> > > 2.26.2
> > >
More information about the igt-dev
mailing list