[PATCH 12/12] drm/i915/xe3lpd: Power request asserting/deasserting
Kahola, Mika
mika.kahola at intel.com
Mon Oct 21 13:03:47 UTC 2024
> -----Original Message-----
> From: Sousa, Gustavo <gustavo.sousa at intel.com>
> Sent: Monday, 21 October 2024 15.31
> To: Atwood, Matthew S <matthew.s.atwood at intel.com>; intel-
> gfx at lists.freedesktop.org; intel-xe at lists.freedesktop.org
> Cc: Kahola, Mika <mika.kahola at intel.com>; Atwood, Matthew S
> <matthew.s.atwood at intel.com>
> Subject: Re: [PATCH 12/12] drm/i915/xe3lpd: Power request
> asserting/deasserting
>
> Quoting Matt Atwood (2024-10-18 17:49:41-03:00)
> >From: Mika Kahola <mika.kahola at intel.com>
> >
> >There is a HW issue that arises when there are race conditions between
> >TCSS entering/exiting TC7 or TC10 states while the driver is
> >asserting/deasserting TCSS power request. As a workaround, Display
> >driver will implement a mailbox sequence to ensure that the TCSS is in
> >TC0 when TCSS power request is asserted/deasserted.
> >
> >The sequence is the following
> >
> >1. Read mailbox command status and wait until run/busy bit is
> > clear
> >2. Write mailbox data value '1' for power request asserting
> > and '0' for power request deasserting 3. Write mailbox command
> >run/busy bit and command value with 0x1 4. Read mailbox command and
> >wait until run/busy bit is clear
> > before continuing power request.
>
> Is there a WA lineage number we can refer to for this?
Unfortunately, there isn't any official WA number for this. This is somewhat unofficially proposed fix.
-Mika-
>
> --
> Gustavo Sousa
>
> >
> >Signed-off-by: Mika Kahola <mika.kahola at intel.com>
> >Signed-off-by: Matt Atwood <matthew.s.atwood at intel.com>
> >---
> > drivers/gpu/drm/i915/display/intel_tc.c | 40 +++++++++++++++++++++++++
> > drivers/gpu/drm/i915/i915_reg.h | 7 +++++
> > 2 files changed, 47 insertions(+)
> >
> >diff --git a/drivers/gpu/drm/i915/display/intel_tc.c
> >b/drivers/gpu/drm/i915/display/intel_tc.c
> >index 6f2ee7dbc43b..7d9f87db381c 100644
> >--- a/drivers/gpu/drm/i915/display/intel_tc.c
> >+++ b/drivers/gpu/drm/i915/display/intel_tc.c
> >@@ -1013,6 +1013,39 @@ xelpdp_tc_phy_wait_for_tcss_power(struct
> intel_tc_port *tc, bool enabled)
> > return true;
> > }
> >
> >+static bool xelpdp_tc_phy_wait_for_tcss_ready(struct drm_i915_private *i915,
> >+ bool enable) {
> >+ if (DISPLAY_VER(i915) < 30)
> >+ return true;
> >+
> >+ /* check if mailbox is running busy */
> >+ if (intel_de_wait_for_clear(i915, TCSS_DISP_MAILBOX_IN_CMD,
> >+ TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY, 10)) {
> >+ drm_dbg_kms(&i915->drm,
> >+ "timeout waiting for TCSS mailbox run/busy bit to clear\n");
> >+ return false;
> >+ }
> >+
> >+ if (enable)
> >+ intel_de_write(i915, TCSS_DISP_MAILBOX_IN_DATA, 1);
> >+ else
> >+ intel_de_write(i915, TCSS_DISP_MAILBOX_IN_DATA, 0);
> >+
> >+ intel_de_write(i915, TCSS_DISP_MAILBOX_IN_CMD,
> >+ TCSS_DISP_MAILBOX_IN_CMD_DATA(1));
> >+
> >+ /* wait to clear mailbox running busy bit before continuing */
> >+ if (intel_de_wait_for_clear(i915, TCSS_DISP_MAILBOX_IN_CMD,
> >+ TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY, 10)) {
> >+ drm_dbg_kms(&i915->drm,
> >+ "timeout waiting for TCSS mailbox run/busy bit to clear\n");
> >+ return false;
> >+ }
> >+
> >+ return true;
> >+}
> >+
> > static void __xelpdp_tc_phy_enable_tcss_power(struct intel_tc_port
> >*tc, bool enable) {
> > struct drm_i915_private *i915 = tc_to_i915(tc); @@ -1022,6
> >+1055,13 @@ static void __xelpdp_tc_phy_enable_tcss_power(struct
> >intel_tc_port *tc, bool ena
> >
> > assert_tc_cold_blocked(tc);
> >
> >+ /*
> >+ * Gfx driver workaround for PTL tcss_rxdetect_clkswb_req/ack handshake
> >+ * violation when pwwreq= 0->1 during TC7/10 entry
> >+ */
> >+ drm_WARN_ON(&i915->drm,
> >+ !xelpdp_tc_phy_wait_for_tcss_ready(i915, enable));
> >+
> > val = intel_de_read(i915, reg);
> > if (enable)
> > val |= XELPDP_TCSS_POWER_REQUEST; diff --git
> >a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
> >index 2743a2dd0a3d..d2775a32bf18 100644
> >--- a/drivers/gpu/drm/i915/i915_reg.h
> >+++ b/drivers/gpu/drm/i915/i915_reg.h
> >@@ -4539,6 +4539,13 @@ enum skl_power_gate {
> > #define TCSS_DDI_STATUS_HPD_LIVE_STATUS_TBT REG_BIT(1)
> > #define TCSS_DDI_STATUS_HPD_LIVE_STATUS_ALT REG_BIT(0)
> >
> >+#define TCSS_DISP_MAILBOX_IN_CMD _MMIO(0x161300)
> >+#define TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY REG_BIT(31)
> >+#define TCSS_DISP_MAILBOX_IN_CMD_CMD_MASK REG_GENMASK(7, 0)
> >+#define TCSS_DISP_MAILBOX_IN_CMD_DATA(x)
> TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY | \
> >+
> REG_FIELD_PREP(TCSS_DISP_MAILBOX_IN_CMD_CMD_MASK, (x))
> >+#define TCSS_DISP_MAILBOX_IN_DATA _MMIO(0x161304)
> >+
> > #define PRIMARY_SPI_TRIGGER _MMIO(0x102040)
> > #define PRIMARY_SPI_ADDRESS _MMIO(0x102080)
> > #define PRIMARY_SPI_REGIONID _MMIO(0x102084)
> >--
> >2.45.0
> >
More information about the Intel-gfx
mailing list