Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [High] drm/panel: uninitialized return value in k101_im2ba02_enable() on DSI error - [Low] drm/panel: single-call mipi_dsi_multi_context misuse and duplicate dev_err() --
commit 98ea9fe2e57e340101561581a07219ed561f36d0 Author: amit <[email protected]> drm/panel: Replace deprecated DCS display-on calls This commit replaces calls to mipi_dsi_dcs_set_display_on() across 14 DRM panel drivers with mipi_dsi_dcs_set_display_on_multi() using a local mipi_dsi_multi_context structure. > diff --git a/drivers/gpu/drm/panel/panel-boe-himax8279d.c > b/drivers/gpu/drm/panel/panel-boe-himax8279d.c > index 4a8560b4b899a..a4b35e042f315 100644 > --- a/drivers/gpu/drm/panel/panel-boe-himax8279d.c > +++ b/drivers/gpu/drm/panel/panel-boe-himax8279d.c > @@ -155,7 +155,12 @@ static int boe_panel_prepare(struct drm_panel *panel) > /* T6: 120ms - 121ms */ > usleep_range(120000, 121000); > > - err = mipi_dsi_dcs_set_display_on(pinfo->link); > + struct mipi_dsi_multi_context dsi_ctx = { > + .dsi = pinfo->link, > + }; > + > + mipi_dsi_dcs_set_display_on_multi(&dsi_ctx); > + err = dsi_ctx.accum_err; > if (err < 0) { > dev_err(panel->dev, "failed to set display on: %d\n", err); > goto poweroff; [Severity: Low] This isn't a bug, but is it intended to instantiate struct mipi_dsi_multi_context mid-function for just a single mipi_dsi_dcs_set_display_on_multi() call while leaving the surrounding DSI calls (such as mipi_dsi_dcs_exit_sleep_mode(), mipi_dsi_dcs_set_tear_on(), mipi_dsi_dcs_set_pixel_format(), and mipi_dsi_dcs_write_buffer()) unconverted? In addition, mipi_dsi_dcs_set_display_on_multi() in drivers/gpu/drm/drm_mipi_dsi.c already logs an error on failure: ret = mipi_dsi_dcs_set_display_on(dsi); if (ret < 0) { ctx->accum_err = ret; dev_err(dev, "sending DCS SET_DISPLAY_ON failed: %d\n", ctx->accum_err); } Keeping the caller's dev_err() here in boe_panel_prepare() (as well as in boe_panel_enable(), innolux_panel_prepare(), khadas_ts050_panel_prepare(), kingdisplay_panel_prepare(), ltk500hd1829_prepare(), panel_nv3051d_prepare(), nt35510_prepare(), rad_panel_enable(), and sharp_panel_prepare()) will print duplicate error messages when setting the display on fails. Declaring dsi_ctx in the middle of the function after executable statements also introduces mixed declarations across 12 of the modified drivers (including dsicm_power_on(), rm68200_prepare(), and s6e63j0x03_enable()). Should the full DSI command sequences in these functions be converted to use a single mipi_dsi_multi_context declared at the top of each function with the redundant dev_err() calls removed? [ ... ] > diff --git a/drivers/gpu/drm/panel/panel-feixin-k101-im2ba02.c > b/drivers/gpu/drm/panel/panel-feixin-k101-im2ba02.c > index 8c3a231c147d5..baa64b7aeb0bd 100644 > --- a/drivers/gpu/drm/panel/panel-feixin-k101-im2ba02.c > +++ b/drivers/gpu/drm/panel/panel-feixin-k101-im2ba02.c > @@ -344,12 +344,15 @@ static int k101_im2ba02_enable(struct drm_panel *panel) > { > struct k101_im2ba02 *ctx = panel_to_k101_im2ba02(panel); > const struct k101_im2ba02_init_cmd *cmd = &timed_cmds[1]; > + struct mipi_dsi_multi_context dsi_ctx = { > + .dsi = ctx->dsi, > + }; > int ret; > > msleep(150); > > - ret = mipi_dsi_dcs_set_display_on(ctx->dsi); > - if (ret < 0) > + mipi_dsi_dcs_set_display_on_multi(&dsi_ctx); > + if (dsi_ctx.accum_err) > return ret; [Severity: High] In k101_im2ba02_enable(), doesn't this return the uninitialized local variable ret when mipi_dsi_dcs_set_display_on_multi() fails and sets dsi_ctx.accum_err? Since ret is never initialized or assigned, returning ret on error triggers an uninitialized variable warning and can return a non-negative garbage value to drm_panel_enable(), causing it to treat a failed panel enable as successful while skipping the final mipi_dsi_dcs_write_buffer() command. Should ret be removed and dsi_ctx.accum_err returned directly instead? -- Sashiko AI review ยท https://sashiko.dev/#/patchset/[email protected]?part=1
