* FAILED: patch "[PATCH] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable" failed to apply to 5.10-stable tree
@ 2026-09-09 11:54 gregkh
2026-09-12 15:08 ` [PATCH 5.10.y 1/3] drm/msm/dsi: do not enable PHYs when called for the slave DSI interface Sasha Levin
0 siblings, 1 reply; 4+ messages in thread
From: gregkh @ 2026-09-09 11:54 UTC (permalink / raw)
To: kavansmith82, daniel, dmitry.baryshkov; +Cc: stable
The patch below does not apply to the 5.10-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable@vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y
git checkout FETCH_HEAD
git cherry-pick -x 6cd33b6f4155efc20485929fd0b56bb704641db9
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable@vger.kernel.org>' --in-reply-to '2026090910-activist-busily-77c5@gregkh' --subject-prefix 'PATCH 5.10.y' 'HEAD^..'
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 6cd33b6f4155efc20485929fd0b56bb704641db9 Mon Sep 17 00:00:00 2001
From: Kavan Smith <kavansmith82@gmail.com>
Date: Mon, 6 Jul 2026 18:32:40 -0700
Subject: [PATCH] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable
value
MSM8916 runtime DSI commands still go through
msm_dsi_host_xfer_prepare(), which re-applies the link clock rate before
enabling the link clocks. That is fine in principle, but on DSI 6G the
requested byte clock rate often does not exactly match the DSI PHY PLL's
realizable rate. For example, the driver can request 56250000 Hz while the
PLL actually runs at 56246337 Hz.
Because the requested and actual rates differ slightly, every later
link_clk_set_rate() call is treated as a real clock change and re-locks
the PLL. On a video-mode panel without an internal timing generator, such
as samsung,s6d7aa0 / lsl080al03 on MSM8916, that live-clock glitch makes
the panel lose pixel lock and visibly corrupts scanout on each runtime DCS
command, including backlight writes.
Fix this by rounding the computed 6G byte clock rate up front, before it is
stored in msm_host->byte_clk_rate and reused by later transfers. Once the
host carries the PLL-achievable rate instead of the idealized one,
repeated link_clk_set_rate() calls become no-ops in the common clock
framework and no longer re-lock the PLL.
This keeps the normal transfer callback sequencing intact, preserves the
OPP vote path in link_clk_set_rate(), and matches the fix direction
suggested in the original 2018 discussion.
Reported-by: Daniel Mack <daniel@zonque.org>
Closes: https://lore.kernel.org/all/1a682c5b-7fc9-3aaa-120b-64b239a355a3@zonque.org/
Fixes: 6b16f05aa39f ("drm/msm/dsi: Split clk rate setting and enable")
Cc: stable@vger.kernel.org
Signed-off-by: Kavan Smith <kavansmith82@gmail.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/738234/
Link: https://lore.kernel.org/r/20260707013240.681012-1-kavansmith82@gmail.com
[DB: dropped extra chunk from the patch]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 982abaaac00d..5e1b313f04c0 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -670,12 +670,24 @@ static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
+ long rounded_byte_clk_rate;
+
if (!msm_host->mode) {
pr_err("%s: mode not set\n", __func__);
return -EINVAL;
}
dsi_calc_pclk(msm_host, is_bonded_dsi);
+
+ rounded_byte_clk_rate = clk_round_rate(msm_host->byte_clk,
+ msm_host->byte_clk_rate);
+ if (rounded_byte_clk_rate < 0) {
+ pr_err("%s: failed to round byte clock rate, %ld\n",
+ __func__, rounded_byte_clk_rate);
+ return rounded_byte_clk_rate;
+ }
+
+ msm_host->byte_clk_rate = rounded_byte_clk_rate;
msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk);
return 0;
}
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.10.y 1/3] drm/msm/dsi: do not enable PHYs when called for the slave DSI interface
2026-09-09 11:54 FAILED: patch "[PATCH] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable" failed to apply to 5.10-stable tree gregkh
@ 2026-09-12 15:08 ` Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 2/3] drm/msm/dsi: rename dual DSI to bonded DSI Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 3/3] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value Sasha Levin
0 siblings, 2 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-12 15:08 UTC (permalink / raw)
To: stable; +Cc: Dmitry Baryshkov, Rob Clark, Sasha Levin
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
[ Upstream commit a14440042fe8dac6b7c507a946f46a6f50c9c05d ]
Move the call to dsi_mgr_phy_enable after checking whether the DSI
interface is slave, so that PHY enablement happens together with the
host enablement.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20210609211211.2561090-1-dmitry.baryshkov@linaro.org
Signed-off-by: Rob Clark <robdclark@chromium.org>
Stable-dep-of: 6cd33b6f4155 ("drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/gpu/drm/msm/dsi/dsi_manager.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c
index fb421ca56b3da..5736665da97b5 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_manager.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c
@@ -381,14 +381,14 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge)
if (!msm_dsi_device_connected(msm_dsi))
return;
- ret = dsi_mgr_phy_enable(id, phy_shared_timings);
- if (ret)
- goto phy_en_fail;
-
/* Do nothing with the host if it is slave-DSI in case of dual DSI */
if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
return;
+ ret = dsi_mgr_phy_enable(id, phy_shared_timings);
+ if (ret)
+ goto phy_en_fail;
+
ret = msm_dsi_host_power_on(host, &phy_shared_timings[id], is_dual_dsi);
if (ret) {
pr_err("%s: power on host %d failed, %d\n", __func__, id, ret);
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.10.y 2/3] drm/msm/dsi: rename dual DSI to bonded DSI
2026-09-12 15:08 ` [PATCH 5.10.y 1/3] drm/msm/dsi: do not enable PHYs when called for the slave DSI interface Sasha Levin
@ 2026-09-12 15:08 ` Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 3/3] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value Sasha Levin
1 sibling, 0 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-12 15:08 UTC (permalink / raw)
To: stable
Cc: Dmitry Baryshkov, Abhinav Kumar, Bjorn Andersson, Rob Clark,
Sasha Levin
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
[ Upstream commit 6183606da324b934166e7b47c3be6176fa86f9ff ]
We are preparing to support two independent DSI hosts in the DSI/DPU
code. To remove possible confusion (as both configurations can be
referenced as dual DSI) let's rename old "dual DSI" (two DSI hosts
driving single device, with clocks being locked) to "bonded DSI".
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Abhinav Kumar <abhinavk@codeaurora.org>
Link: https://lore.kernel.org/r/20210717124016.316020-2-dmitry.baryshkov@linaro.org
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
[DB: add one extra hunk added by one previous patches]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
[5.10 dependency adaptation: keep the existing PHY/PLL interfaces and
rename the stable-only source PLL selection to IS_BONDED_DSI(). Apply
the 10nm, 14nm and 7nm PLL comment changes in dsi/pll/, where that code
lives in this tree. Omit the test-pattern hunk and newer host declarations,
which have no implementation here. No functions or clock behavior are
added. The resulting host names let the target byte-clock rounding patch
apply without further changes.]
[ sashal: Reduced backport -- upstream 6183606da324b touches 8 file(s), this
backport carries 8. Not backported here:
drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c
drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c
drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c
This note is generated from the file lists only; see the resolution record
for the reasoning. ]
Stable-dep-of: 6cd33b6f4155 ("drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/gpu/drm/msm/disp/mdp5/mdp5_ctl.c | 2 +-
drivers/gpu/drm/msm/dsi/dsi.h | 8 +-
drivers/gpu/drm/msm/dsi/dsi_cfg.h | 2 +-
drivers/gpu/drm/msm/dsi/dsi_host.c | 34 +++----
drivers/gpu/drm/msm/dsi/dsi_manager.c | 103 ++++++++++-----------
drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c | 2 +-
drivers/gpu/drm/msm/dsi/pll/dsi_pll_14nm.c | 6 +-
drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c | 2 +-
8 files changed, 79 insertions(+), 80 deletions(-)
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_ctl.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_ctl.c
index 030279d7b64b7..9399055518a49 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_ctl.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_ctl.c
@@ -728,7 +728,7 @@ struct mdp5_ctl_manager *mdp5_ctlm_init(struct drm_device *dev,
}
/*
- * In Dual DSI case, CTL0 and CTL1 are always assigned to two DSI
+ * In bonded DSI case, CTL0 and CTL1 are always assigned to two DSI
* interfaces to support single FLUSH feature (Flush CTL0 and CTL1 when
* only write into CTL0's FLUSH register) to keep two DSI pipes in sync.
* Single FLUSH is supported from hw rev v3.0.
diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h
index 78ef5d4ed922f..f3c24ac7b9381 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.h
@@ -159,7 +159,7 @@ int msm_dsi_host_enable(struct mipi_dsi_host *host);
int msm_dsi_host_disable(struct mipi_dsi_host *host);
int msm_dsi_host_power_on(struct mipi_dsi_host *host,
struct msm_dsi_phy_shared_timings *phy_shared_timings,
- bool is_dual_dsi);
+ bool is_bonded_dsi);
int msm_dsi_host_power_off(struct mipi_dsi_host *host);
int msm_dsi_host_set_display_mode(struct mipi_dsi_host *host,
const struct drm_display_mode *mode);
@@ -173,7 +173,7 @@ int msm_dsi_host_set_src_pll(struct mipi_dsi_host *host,
void msm_dsi_host_reset_phy(struct mipi_dsi_host *host);
void msm_dsi_host_get_phy_clk_req(struct mipi_dsi_host *host,
struct msm_dsi_phy_clk_request *clk_req,
- bool is_dual_dsi);
+ bool is_bonded_dsi);
void msm_dsi_host_destroy(struct mipi_dsi_host *host);
int msm_dsi_host_modeset_init(struct mipi_dsi_host *host,
struct drm_device *dev);
@@ -195,8 +195,8 @@ int dsi_dma_base_get_6g(struct msm_dsi_host *msm_host, uint64_t *iova);
int dsi_dma_base_get_v2(struct msm_dsi_host *msm_host, uint64_t *iova);
int dsi_clk_init_v2(struct msm_dsi_host *msm_host);
int dsi_clk_init_6g_v2(struct msm_dsi_host *msm_host);
-int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_dual_dsi);
-int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_dual_dsi);
+int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_bonded_dsi);
+int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_bonded_dsi);
/* dsi phy */
struct msm_dsi_phy;
diff --git a/drivers/gpu/drm/msm/dsi/dsi_cfg.h b/drivers/gpu/drm/msm/dsi/dsi_cfg.h
index 89386b10dc48b..169625c473fe9 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_cfg.h
+++ b/drivers/gpu/drm/msm/dsi/dsi_cfg.h
@@ -47,7 +47,7 @@ struct msm_dsi_host_cfg_ops {
void* (*tx_buf_get)(struct msm_dsi_host *msm_host);
void (*tx_buf_put)(struct msm_dsi_host *msm_host);
int (*dma_base_get)(struct msm_dsi_host *msm_host, uint64_t *iova);
- int (*calc_clk_rate)(struct msm_dsi_host *msm_host, bool is_dual_dsi);
+ int (*calc_clk_rate)(struct msm_dsi_host *msm_host, bool is_bonded_dsi);
};
struct msm_dsi_cfg_handler {
diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index b69099b533bfe..8066a02abe8d1 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -680,7 +680,7 @@ void dsi_link_clk_disable_v2(struct msm_dsi_host *msm_host)
clk_disable_unprepare(msm_host->byte_clk);
}
-static u32 dsi_get_pclk_rate(struct msm_dsi_host *msm_host, bool is_dual_dsi)
+static u32 dsi_get_pclk_rate(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
struct drm_display_mode *mode = msm_host->mode;
u32 pclk_rate;
@@ -688,22 +688,22 @@ static u32 dsi_get_pclk_rate(struct msm_dsi_host *msm_host, bool is_dual_dsi)
pclk_rate = mode->clock * 1000u;
/*
- * For dual DSI mode, the current DRM mode has the complete width of the
+ * For bonded DSI mode, the current DRM mode has the complete width of the
* panel. Since, the complete panel is driven by two DSI controllers,
* the clock rates have to be split between the two dsi controllers.
* Adjust the byte and pixel clock rates for each dsi host accordingly.
*/
- if (is_dual_dsi)
+ if (is_bonded_dsi)
pclk_rate /= 2;
return pclk_rate;
}
-static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_dual_dsi)
+static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
u8 lanes = msm_host->lanes;
u32 bpp = dsi_get_bpp(msm_host->format);
- u32 pclk_rate = dsi_get_pclk_rate(msm_host, is_dual_dsi);
+ u32 pclk_rate = dsi_get_pclk_rate(msm_host, is_bonded_dsi);
u64 pclk_bpp = (u64)pclk_rate * bpp;
if (lanes == 0) {
@@ -721,28 +721,28 @@ static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_dual_dsi)
}
-int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_dual_dsi)
+int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
if (!msm_host->mode) {
pr_err("%s: mode not set\n", __func__);
return -EINVAL;
}
- dsi_calc_pclk(msm_host, is_dual_dsi);
+ dsi_calc_pclk(msm_host, is_bonded_dsi);
msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk);
return 0;
}
-int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_dual_dsi)
+int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
u32 bpp = dsi_get_bpp(msm_host->format);
u64 pclk_bpp;
unsigned int esc_mhz, esc_div;
unsigned long byte_mhz;
- dsi_calc_pclk(msm_host, is_dual_dsi);
+ dsi_calc_pclk(msm_host, is_bonded_dsi);
- pclk_bpp = (u64)dsi_get_pclk_rate(msm_host, is_dual_dsi) * bpp;
+ pclk_bpp = (u64)dsi_get_pclk_rate(msm_host, is_bonded_dsi) * bpp;
do_div(pclk_bpp, 8);
msm_host->src_clk_rate = pclk_bpp;
@@ -939,7 +939,7 @@ static void dsi_ctrl_config(struct msm_dsi_host *msm_host, bool enable,
dsi_write(msm_host, REG_DSI_CTRL, data);
}
-static void dsi_timing_setup(struct msm_dsi_host *msm_host, bool is_dual_dsi)
+static void dsi_timing_setup(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
struct drm_display_mode *mode = msm_host->mode;
u32 hs_start = 0, vs_start = 0; /* take sync start as 0 */
@@ -957,13 +957,13 @@ static void dsi_timing_setup(struct msm_dsi_host *msm_host, bool is_dual_dsi)
DBG("");
/*
- * For dual DSI mode, the current DRM mode has
+ * For bonded DSI mode, the current DRM mode has
* the complete width of the panel. Since, the complete
* panel is driven by two DSI controllers, the horizontal
* timings have to be split between the two dsi controllers.
* Adjust the DSI host timing values accordingly.
*/
- if (is_dual_dsi) {
+ if (is_bonded_dsi) {
h_total /= 2;
hs_end /= 2;
ha_start /= 2;
@@ -2318,13 +2318,13 @@ void msm_dsi_host_reset_phy(struct mipi_dsi_host *host)
void msm_dsi_host_get_phy_clk_req(struct mipi_dsi_host *host,
struct msm_dsi_phy_clk_request *clk_req,
- bool is_dual_dsi)
+ bool is_bonded_dsi)
{
struct msm_dsi_host *msm_host = to_msm_dsi_host(host);
const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd;
int ret;
- ret = cfg_hnd->ops->calc_clk_rate(msm_host, is_dual_dsi);
+ ret = cfg_hnd->ops->calc_clk_rate(msm_host, is_bonded_dsi);
if (ret) {
pr_err("%s: unable to calc clk rate, %d\n", __func__, ret);
return;
@@ -2387,7 +2387,7 @@ static void msm_dsi_sfpb_config(struct msm_dsi_host *msm_host, bool enable)
int msm_dsi_host_power_on(struct mipi_dsi_host *host,
struct msm_dsi_phy_shared_timings *phy_shared_timings,
- bool is_dual_dsi)
+ bool is_bonded_dsi)
{
struct msm_dsi_host *msm_host = to_msm_dsi_host(host);
const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd;
@@ -2425,7 +2425,7 @@ int msm_dsi_host_power_on(struct mipi_dsi_host *host,
goto fail_disable_clk;
}
- dsi_timing_setup(msm_host, is_dual_dsi);
+ dsi_timing_setup(msm_host, is_bonded_dsi);
dsi_sw_reset(msm_host);
dsi_ctrl_config(msm_host, true, phy_shared_timings);
diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c
index 5736665da97b5..0197304539622 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_manager.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c
@@ -21,14 +21,14 @@
struct msm_dsi_manager {
struct msm_dsi *dsi[DSI_MAX];
- bool is_dual_dsi;
+ bool is_bonded_dsi;
bool is_sync_needed;
int master_dsi_link_id;
};
static struct msm_dsi_manager msm_dsim_glb;
-#define IS_DUAL_DSI() (msm_dsim_glb.is_dual_dsi)
+#define IS_BONDED_DSI() (msm_dsim_glb.is_bonded_dsi)
#define IS_SYNC_NEEDED() (msm_dsim_glb.is_sync_needed)
#define IS_MASTER_DSI_LINK(id) (msm_dsim_glb.master_dsi_link_id == id)
@@ -42,18 +42,17 @@ static inline struct msm_dsi *dsi_mgr_get_other_dsi(int id)
return msm_dsim_glb.dsi[(id + 1) % DSI_MAX];
}
-static int dsi_mgr_parse_dual_dsi(struct device_node *np, int id)
+static int dsi_mgr_parse_of(struct device_node *np, int id)
{
struct msm_dsi_manager *msm_dsim = &msm_dsim_glb;
- /* We assume 2 dsi nodes have the same information of dual-dsi and
- * sync-mode, and only one node specifies master in case of dual mode.
+ /* We assume 2 dsi nodes have the same information of bonded dsi and
+ * sync-mode, and only one node specifies master in case of bonded mode.
*/
- if (!msm_dsim->is_dual_dsi)
- msm_dsim->is_dual_dsi = of_property_read_bool(
- np, "qcom,dual-dsi-mode");
+ if (!msm_dsim->is_bonded_dsi)
+ msm_dsim->is_bonded_dsi = of_property_read_bool(np, "qcom,dual-dsi-mode");
- if (msm_dsim->is_dual_dsi) {
+ if (msm_dsim->is_bonded_dsi) {
if (of_property_read_bool(np, "qcom,master-dsi"))
msm_dsim->master_dsi_link_id = id;
if (!msm_dsim->is_sync_needed)
@@ -73,7 +72,7 @@ static int dsi_mgr_setup_components(int id)
struct msm_dsi_pll *src_pll;
int ret;
- if (!IS_DUAL_DSI()) {
+ if (!IS_BONDED_DSI()) {
ret = msm_dsi_host_register(msm_dsi->host, true);
if (ret)
return ret;
@@ -104,7 +103,7 @@ static int dsi_mgr_setup_components(int id)
if (ret)
return ret;
- /* PLL0 is to drive both 2 DSI link clocks in Dual DSI mode. */
+ /* PLL0 is to drive both 2 DSI link clocks in bonded DSI mode. */
msm_dsi_phy_set_usecase(clk_master_dsi->phy,
MSM_DSI_PHY_MASTER);
msm_dsi_phy_set_usecase(clk_slave_dsi->phy,
@@ -126,9 +125,9 @@ static int enable_phy(struct msm_dsi *msm_dsi, int src_pll_id,
{
struct msm_dsi_phy_clk_request clk_req;
int ret;
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
- msm_dsi_host_get_phy_clk_req(msm_dsi->host, &clk_req, is_dual_dsi);
+ msm_dsi_host_get_phy_clk_req(msm_dsi->host, &clk_req, is_bonded_dsi);
ret = msm_dsi_phy_enable(msm_dsi->phy, src_pll_id, &clk_req);
msm_dsi_phy_get_shared_timings(msm_dsi->phy, shared_timings);
@@ -143,15 +142,15 @@ dsi_mgr_phy_enable(int id,
struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id);
struct msm_dsi *mdsi = dsi_mgr_get_dsi(DSI_CLOCK_MASTER);
struct msm_dsi *sdsi = dsi_mgr_get_dsi(DSI_CLOCK_SLAVE);
- int src_pll_id = IS_DUAL_DSI() ? DSI_CLOCK_MASTER : id;
+ int src_pll_id = IS_BONDED_DSI() ? DSI_CLOCK_MASTER : id;
int ret;
- /* In case of dual DSI, some registers in PHY1 have been programmed
+ /* In case of bonded DSI, some registers in PHY1 have been programmed
* during PLL0 clock's set_rate. The PHY1 reset called by host1 here
* will silently reset those PHY1 registers. Therefore we need to reset
* and enable both PHYs before any PLL clock operation.
*/
- if (IS_DUAL_DSI() && mdsi && sdsi) {
+ if (IS_BONDED_DSI() && mdsi && sdsi) {
if (!mdsi->phy_enabled && !sdsi->phy_enabled) {
msm_dsi_host_reset_phy(mdsi->host);
msm_dsi_host_reset_phy(sdsi->host);
@@ -186,11 +185,11 @@ static void dsi_mgr_phy_disable(int id)
struct msm_dsi *sdsi = dsi_mgr_get_dsi(DSI_CLOCK_SLAVE);
/* disable DSI phy
- * In dual-dsi configuration, the phy should be disabled for the
+ * In bonded dsi configuration, the phy should be disabled for the
* first controller only when the second controller is disabled.
*/
msm_dsi->phy_enabled = false;
- if (IS_DUAL_DSI() && mdsi && sdsi) {
+ if (IS_BONDED_DSI() && mdsi && sdsi) {
if (!mdsi->phy_enabled && !sdsi->phy_enabled) {
msm_dsi_phy_disable(sdsi->phy);
msm_dsi_phy_disable(mdsi->phy);
@@ -252,7 +251,7 @@ static int msm_dsi_manager_panel_init(struct drm_connector *conn, u8 id)
struct msm_dsi *master_dsi, *slave_dsi;
struct drm_panel *panel;
- if (IS_DUAL_DSI() && !IS_MASTER_DSI_LINK(id)) {
+ if (IS_BONDED_DSI() && !IS_MASTER_DSI_LINK(id)) {
master_dsi = other_dsi;
slave_dsi = msm_dsi;
} else {
@@ -261,7 +260,7 @@ static int msm_dsi_manager_panel_init(struct drm_connector *conn, u8 id)
}
/*
- * There is only 1 panel in the global panel list for dual DSI mode.
+ * There is only 1 panel in the global panel list for bonded DSI mode.
* Therefore slave dsi should get the drm_panel instance from master
* dsi.
*/
@@ -272,14 +271,14 @@ static int msm_dsi_manager_panel_init(struct drm_connector *conn, u8 id)
return PTR_ERR(panel);
}
- if (!panel || !IS_DUAL_DSI())
+ if (!panel || !IS_BONDED_DSI())
goto out;
drm_object_attach_property(&conn->base,
conn->dev->mode_config.tile_property, 0);
/*
- * Set split display info to kms once dual DSI panel is connected to
+ * Set split display info to kms once bonded DSI panel is connected to
* both hosts.
*/
if (other_dsi && other_dsi->panel && kms->funcs->set_split_display) {
@@ -325,7 +324,7 @@ static int dsi_mgr_connector_get_modes(struct drm_connector *connector)
return 0;
/*
- * In dual DSI mode, we have one connector that can be
+ * In bonded DSI mode, we have one connector that can be
* attached to the drm_panel.
*/
num = drm_panel_get_modes(panel, connector);
@@ -374,30 +373,30 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge)
struct mipi_dsi_host *host = msm_dsi->host;
struct drm_panel *panel = msm_dsi->panel;
struct msm_dsi_phy_shared_timings phy_shared_timings[DSI_MAX];
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
int ret;
DBG("id=%d", id);
if (!msm_dsi_device_connected(msm_dsi))
return;
- /* Do nothing with the host if it is slave-DSI in case of dual DSI */
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
+ /* Do nothing with the host if it is slave-DSI in case of bonded DSI */
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id))
return;
ret = dsi_mgr_phy_enable(id, phy_shared_timings);
if (ret)
goto phy_en_fail;
- ret = msm_dsi_host_power_on(host, &phy_shared_timings[id], is_dual_dsi);
+ ret = msm_dsi_host_power_on(host, &phy_shared_timings[id], is_bonded_dsi);
if (ret) {
pr_err("%s: power on host %d failed, %d\n", __func__, id, ret);
goto host_on_fail;
}
- if (is_dual_dsi && msm_dsi1) {
+ if (is_bonded_dsi && msm_dsi1) {
ret = msm_dsi_host_power_on(msm_dsi1->host,
- &phy_shared_timings[DSI_1], is_dual_dsi);
+ &phy_shared_timings[DSI_1], is_bonded_dsi);
if (ret) {
pr_err("%s: power on host1 failed, %d\n",
__func__, ret);
@@ -423,7 +422,7 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge)
goto host_en_fail;
}
- if (is_dual_dsi && msm_dsi1) {
+ if (is_bonded_dsi && msm_dsi1) {
ret = msm_dsi_host_enable(msm_dsi1->host);
if (ret) {
pr_err("%s: enable host1 failed, %d\n", __func__, ret);
@@ -439,7 +438,7 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge)
if (panel)
drm_panel_unprepare(panel);
panel_prep_fail:
- if (is_dual_dsi && msm_dsi1)
+ if (is_bonded_dsi && msm_dsi1)
msm_dsi_host_power_off(msm_dsi1->host);
host1_on_fail:
msm_dsi_host_power_off(host);
@@ -454,15 +453,15 @@ static void dsi_mgr_bridge_enable(struct drm_bridge *bridge)
int id = dsi_mgr_bridge_get_id(bridge);
struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id);
struct drm_panel *panel = msm_dsi->panel;
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
int ret;
DBG("id=%d", id);
if (!msm_dsi_device_connected(msm_dsi))
return;
- /* Do nothing with the host if it is slave-DSI in case of dual DSI */
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
+ /* Do nothing with the host if it is slave-DSI in case of bonded DSI */
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id))
return;
if (panel) {
@@ -479,15 +478,15 @@ static void dsi_mgr_bridge_disable(struct drm_bridge *bridge)
int id = dsi_mgr_bridge_get_id(bridge);
struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id);
struct drm_panel *panel = msm_dsi->panel;
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
int ret;
DBG("id=%d", id);
if (!msm_dsi_device_connected(msm_dsi))
return;
- /* Do nothing with the host if it is slave-DSI in case of dual DSI */
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
+ /* Do nothing with the host if it is slave-DSI in case of bonded DSI */
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id))
return;
if (panel) {
@@ -506,7 +505,7 @@ static void dsi_mgr_bridge_post_disable(struct drm_bridge *bridge)
struct mipi_dsi_host *host = msm_dsi->host;
struct drm_panel *panel = msm_dsi->panel;
struct msm_dsi_pll *src_pll;
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
int ret;
DBG("id=%d", id);
@@ -515,18 +514,18 @@ static void dsi_mgr_bridge_post_disable(struct drm_bridge *bridge)
return;
/*
- * Do nothing with the host if it is slave-DSI in case of dual DSI.
+ * Do nothing with the host if it is slave-DSI in case of bonded DSI.
* It is safe to call dsi_mgr_phy_disable() here because a single PHY
* won't be diabled until both PHYs request disable.
*/
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id))
goto disable_phy;
ret = msm_dsi_host_disable(host);
if (ret)
pr_err("%s: host %d disable failed, %d\n", __func__, id, ret);
- if (is_dual_dsi && msm_dsi1) {
+ if (is_bonded_dsi && msm_dsi1) {
ret = msm_dsi_host_disable(msm_dsi1->host);
if (ret)
pr_err("%s: host1 disable failed, %d\n", __func__, ret);
@@ -547,7 +546,7 @@ static void dsi_mgr_bridge_post_disable(struct drm_bridge *bridge)
if (ret)
pr_err("%s: host %d power off failed,%d\n", __func__, id, ret);
- if (is_dual_dsi && msm_dsi1) {
+ if (is_bonded_dsi && msm_dsi1) {
ret = msm_dsi_host_power_off(msm_dsi1->host);
if (ret)
pr_err("%s: host1 power off failed, %d\n",
@@ -566,15 +565,15 @@ static void dsi_mgr_bridge_mode_set(struct drm_bridge *bridge,
struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id);
struct msm_dsi *other_dsi = dsi_mgr_get_other_dsi(id);
struct mipi_dsi_host *host = msm_dsi->host;
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
DBG("set mode: " DRM_MODE_FMT, DRM_MODE_ARG(mode));
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id))
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id))
return;
msm_dsi_host_set_display_mode(host, adjusted_mode);
- if (is_dual_dsi && other_dsi)
+ if (is_bonded_dsi && other_dsi)
msm_dsi_host_set_display_mode(other_dsi->host, adjusted_mode);
}
@@ -650,15 +649,15 @@ struct drm_connector *msm_dsi_manager_connector_init(u8 id)
bool msm_dsi_manager_validate_current_config(u8 id)
{
- bool is_dual_dsi = IS_DUAL_DSI();
+ bool is_bonded_dsi = IS_BONDED_DSI();
/*
- * For dual DSI, we only have one drm panel. For this
+ * For bonded DSI, we only have one drm panel. For this
* use case, we register only one bridge/connector.
* Skip bridge/connector initialisation if it is
- * slave-DSI for dual DSI configuration.
+ * slave-DSI for bonded DSI configuration.
*/
- if (is_dual_dsi && !IS_MASTER_DSI_LINK(id)) {
+ if (is_bonded_dsi && !IS_MASTER_DSI_LINK(id)) {
DBG("Skip bridge registration for slave DSI->id: %d\n", id);
return false;
}
@@ -750,7 +749,7 @@ int msm_dsi_manager_cmd_xfer(int id, const struct mipi_dsi_msg *msg)
if (!msg->tx_buf || !msg->tx_len)
return 0;
- /* In dual master case, panel requires the same commands sent to
+ /* In bonded master case, panel requires the same commands sent to
* both DSI links. Host issues the command trigger to both links
* when DSI_1 calls the cmd transfer function, no matter it happens
* before or after DSI_0 cmd transfer.
@@ -819,9 +818,9 @@ int msm_dsi_manager_register(struct msm_dsi *msm_dsi)
msm_dsim->dsi[id] = msm_dsi;
- ret = dsi_mgr_parse_dual_dsi(msm_dsi->pdev->dev.of_node, id);
+ ret = dsi_mgr_parse_of(msm_dsi->pdev->dev.of_node, id);
if (ret) {
- pr_err("%s: failed to parse dual DSI info\n", __func__);
+ pr_err("%s: failed to parse OF DSI info\n", __func__);
goto fail;
}
diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c
index e4e9bf04b7368..5e0eb94eb20d4 100644
--- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c
+++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c
@@ -124,7 +124,7 @@ struct dsi_pll_10nm {
#define to_pll_10nm(x) container_of(x, struct dsi_pll_10nm, base)
/*
- * Global list of private DSI PLL struct pointers. We need this for Dual DSI
+ * Global list of private DSI PLL struct pointers. We need this for bonded DSI
* mode, where the master PLL's clk_ops needs access the slave's private data
*/
static struct dsi_pll_10nm *pll_10nm_list[DSI_MAX];
diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_14nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_14nm.c
index f847376d501e7..0cd304740b09a 100644
--- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_14nm.c
+++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_14nm.c
@@ -152,7 +152,7 @@ struct dsi_pll_14nm {
/*
* Private struct for N1/N2 post-divider clocks. These clocks are similar to
* the generic clk_divider class of clocks. The only difference is that it
- * also sets the slave DSI PLL's post-dividers if in Dual DSI mode
+ * also sets the slave DSI PLL's post-dividers if in bonded DSI mode
*/
struct dsi_pll_14nm_postdiv {
struct clk_hw hw;
@@ -168,7 +168,7 @@ struct dsi_pll_14nm_postdiv {
#define to_pll_14nm_postdiv(_hw) container_of(_hw, struct dsi_pll_14nm_postdiv, hw)
/*
- * Global list of private DSI PLL struct pointers. We need this for Dual DSI
+ * Global list of private DSI PLL struct pointers. We need this for bonded DSI
* mode, where the master PLL's clk_ops needs access the slave's private data
*/
static struct dsi_pll_14nm *pll_14nm_list[DSI_MAX];
@@ -734,7 +734,7 @@ static int dsi_pll_14nm_postdiv_set_rate(struct clk_hw *hw, unsigned long rate,
val |= value << shift;
pll_write(base + REG_DSI_14nm_PHY_CMN_CLK_CFG0, val);
- /* If we're master in dual DSI mode, then the slave PLL's post-dividers
+ /* If we're master in bonded DSI mode, then the slave PLL's post-dividers
* follow the master's post dividers
*/
if (pll_14nm->uc == MSM_DSI_PHY_MASTER) {
diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c
index c1c41846b6b2b..01438a2520914 100644
--- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c
+++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c
@@ -124,7 +124,7 @@ struct dsi_pll_7nm {
#define to_pll_7nm(x) container_of(x, struct dsi_pll_7nm, base)
/*
- * Global list of private DSI PLL struct pointers. We need this for Dual DSI
+ * Global list of private DSI PLL struct pointers. We need this for bonded DSI
* mode, where the master PLL's clk_ops needs access the slave's private data
*/
static struct dsi_pll_7nm *pll_7nm_list[DSI_MAX];
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.10.y 3/3] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value
2026-09-12 15:08 ` [PATCH 5.10.y 1/3] drm/msm/dsi: do not enable PHYs when called for the slave DSI interface Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 2/3] drm/msm/dsi: rename dual DSI to bonded DSI Sasha Levin
@ 2026-09-12 15:08 ` Sasha Levin
1 sibling, 0 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-12 15:08 UTC (permalink / raw)
To: stable; +Cc: Kavan Smith, Daniel Mack, Dmitry Baryshkov, Sasha Levin
From: Kavan Smith <kavansmith82@gmail.com>
[ Upstream commit 6cd33b6f4155efc20485929fd0b56bb704641db9 ]
MSM8916 runtime DSI commands still go through
msm_dsi_host_xfer_prepare(), which re-applies the link clock rate before
enabling the link clocks. That is fine in principle, but on DSI 6G the
requested byte clock rate often does not exactly match the DSI PHY PLL's
realizable rate. For example, the driver can request 56250000 Hz while the
PLL actually runs at 56246337 Hz.
Because the requested and actual rates differ slightly, every later
link_clk_set_rate() call is treated as a real clock change and re-locks
the PLL. On a video-mode panel without an internal timing generator, such
as samsung,s6d7aa0 / lsl080al03 on MSM8916, that live-clock glitch makes
the panel lose pixel lock and visibly corrupts scanout on each runtime DCS
command, including backlight writes.
Fix this by rounding the computed 6G byte clock rate up front, before it is
stored in msm_host->byte_clk_rate and reused by later transfers. Once the
host carries the PLL-achievable rate instead of the idealized one,
repeated link_clk_set_rate() calls become no-ops in the common clock
framework and no longer re-lock the PLL.
This keeps the normal transfer callback sequencing intact, preserves the
OPP vote path in link_clk_set_rate(), and matches the fix direction
suggested in the original 2018 discussion.
Reported-by: Daniel Mack <daniel@zonque.org>
Closes: https://lore.kernel.org/all/1a682c5b-7fc9-3aaa-120b-64b239a355a3@zonque.org/
Fixes: 6b16f05aa39f ("drm/msm/dsi: Split clk rate setting and enable")
Cc: stable@vger.kernel.org
Signed-off-by: Kavan Smith <kavansmith82@gmail.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/738234/
Link: https://lore.kernel.org/r/20260707013240.681012-1-kavansmith82@gmail.com
[DB: dropped extra chunk from the patch]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/gpu/drm/msm/dsi/dsi_host.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 8066a02abe8d1..4b0998f13d67a 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -723,12 +723,24 @@ static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_bonded_dsi)
{
+ long rounded_byte_clk_rate;
+
if (!msm_host->mode) {
pr_err("%s: mode not set\n", __func__);
return -EINVAL;
}
dsi_calc_pclk(msm_host, is_bonded_dsi);
+
+ rounded_byte_clk_rate = clk_round_rate(msm_host->byte_clk,
+ msm_host->byte_clk_rate);
+ if (rounded_byte_clk_rate < 0) {
+ pr_err("%s: failed to round byte clock rate, %ld\n",
+ __func__, rounded_byte_clk_rate);
+ return rounded_byte_clk_rate;
+ }
+
+ msm_host->byte_clk_rate = rounded_byte_clk_rate;
msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk);
return 0;
}
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-09-12 15:08 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-09-09 11:54 FAILED: patch "[PATCH] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable" failed to apply to 5.10-stable tree gregkh
2026-09-12 15:08 ` [PATCH 5.10.y 1/3] drm/msm/dsi: do not enable PHYs when called for the slave DSI interface Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 2/3] drm/msm/dsi: rename dual DSI to bonded DSI Sasha Levin
2026-09-12 15:08 ` [PATCH 5.10.y 3/3] drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value Sasha Levin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.