Skip to contentSkip to conversation

remoteproc: mediatek: Fix side effect of mt8195 sram power on

PATCH Superseded Tinghan Shen 3 participants 5 messages Mar 16, 2022 – Mar 18, 2022

Subsystems (2)

v4 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

PATCH v4 1/1 remoteproc: mediatek: Fix side effect of mt8195 sram power on

The definition of L1TCM_SRAM_PDN bits on mt8195 is different to mt8192.

L1TCM_SRAM_PDN bits[3:0] control the power of mt8195 L1TCM SRAM.

L1TCM_SRAM_PDN bits[7:4] control the access path to EMI for SCP. These bits have to be powered on to allow EMI access for SCP.

Bits[7:4] also affect audio DSP because audio DSP and SCP are placed on the same hardware bus. If SCP cannot access EMI, audio DSP is blocked too.

L1TCM_SRAM_PDN bits[31:8] are not used.

This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM. It's because the modification introduces a short period of time blocking audio DSP to access EMI. This was not a problem until we have to load both SCP module and audio DSP module. audio DSP needs to access EMI because it has source/data on DRAM. Audio DSP will have unexpected behavior when it accesses EMI and the SCP driver blocks the EMI path at the same time.

Fixes 79111df414fc ("remoteproc: mediatek: Support mt8195 scp")

---
v4: add Fixes and Reviewed-by tags
v3: fix build error
v2: apply comments about macro definition and function calls

---
 drivers/remoteproc/mtk_common.h |  2 ++
 drivers/remoteproc/mtk_scp.c    | 67 +++++++++++++++++++++++++++++++----------
 2 files changed, 53 insertions(+), 16 deletions(-)
drivers/remoteproc/mtk_common.h +2 -0
index 5ff3867c72f3..ff954a06637c 100644
@@ -51,6 +51,8 @@
 #define MT8192_CORE0_WDT_IRQ		0x10030
 #define MT8192_CORE0_WDT_CFG		0x10034
 
+#define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS		GENMASK(7, 4)
+
 #define SCP_FW_VER_LEN			32
 #define SCP_SHARE_BUFFER_SIZE		288
 
drivers/remoteproc/mtk_scp.c +51 -16

106 lines not on this page · raw message

Signature · 1 line
2.15.GIT
maintainer
#1 lore raw

Good morning,

Quoted 52 lines · Tinghan Shen

The definition of L1TCM_SRAM_PDN bits on mt8195 is different to mt8192.

L1TCM_SRAM_PDN bits[3:0] control the power of mt8195 L1TCM SRAM.

L1TCM_SRAM_PDN bits[7:4] control the access path to EMI for SCP. These bits have to be powered on to allow EMI access for SCP.

Bits[7:4] also affect audio DSP because audio DSP and SCP are placed on the same hardware bus. If SCP cannot access EMI, audio DSP is blocked too.

L1TCM_SRAM_PDN bits[31:8] are not used.

This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM. It's because the modification introduces a short period of time blocking audio DSP to access EMI. This was not a problem until we have to load both SCP module and audio DSP module. audio DSP needs to access EMI because it has source/data on DRAM. Audio DSP will have unexpected behavior when it accesses EMI and the SCP driver blocks the EMI path at the same time.

Fixes 79111df414fc ("remoteproc: mediatek: Support mt8195 scp")

---
v4: add Fixes and Reviewed-by tags
v3: fix build error
v2: apply comments about macro definition and function calls

---
 drivers/remoteproc/mtk_common.h |  2 ++
 drivers/remoteproc/mtk_scp.c    | 67 +++++++++++++++++++++++++++++++----------
 2 files changed, 53 insertions(+), 16 deletions(-)
drivers/remoteproc/mtk_common.h +2 -0
index 5ff3867c72f3..ff954a06637c 100644
@@ -51,6 +51,8 @@
 #define MT8192_CORE0_WDT_IRQ		0x10030
 #define MT8192_CORE0_WDT_CFG		0x10034
 
+#define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS		GENMASK(7, 4)
+
 #define SCP_FW_VER_LEN			32
 #define SCP_SHARE_BUFFER_SIZE		288
 
drivers/remoteproc/mtk_scp.c +0 -0
index 36e48cf58ed6..5f686fe09203 100644

drivers/remoteproc/mtk_scp.c static int mt8183_scp_before_load(struct mtk_scp *scp) · line 368

Quoted 5 lines
drivers/remoteproc/mtk_scp.c +1 -1
@@ -365,22 +365,22 @@ static int mt8183_scp_before_load(struct mtk_scp *scp)
 	return 0;
 }
 
-static void mt8192_power_on_sram(void __iomem *addr)
+static void scp_sram_power_on(void __iomem *addr, u32 reserved_mask)

Why is @reserved_mask needed? It is not described in the changelong and as far as I can see in this patchset the parameter is always set to '0', which has no effect on the mask that gets generated.

Thanks,
Mathieu

Quoted 107 lines
drivers/remoteproc/mtk_scp.c +50 -15
 {
 	int i;
 
 	for (i = 31; i >= 0; i--)
-		writel(GENMASK(i, 0), addr);
+		writel(GENMASK(i, 0) & ~reserved_mask, addr);
 	writel(0, addr);
 }
 
-static void mt8192_power_off_sram(void __iomem *addr)
+static void scp_sram_power_off(void __iomem *addr, u32 reserved_mask)
 {
 	int i;
 
 	writel(0, addr);
 	for (i = 0; i < 32; i++)
-		writel(GENMASK(i, 0), addr);
+		writel(GENMASK(i, 0) & ~reserved_mask, addr);
 }
 
 static int mt8192_scp_before_load(struct mtk_scp *scp)
@@ -391,11 +391,32 @@ static int mt8192_scp_before_load(struct mtk_scp *scp)
 	writel(1, scp->reg_base + MT8192_CORE0_SW_RSTN_SET);
 
 	/* enable SRAM clock */
-	mt8192_power_on_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_0);
-	mt8192_power_on_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_1);
-	mt8192_power_on_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_2);
-	mt8192_power_on_sram(scp->reg_base + MT8192_L1TCM_SRAM_PDN);
-	mt8192_power_on_sram(scp->reg_base + MT8192_CPU0_SRAM_PD);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L1TCM_SRAM_PDN, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
+
+	/* enable MPU for all memory regions */
+	writel(0xff, scp->reg_base + MT8192_CORE0_MEM_ATT_PREDEF);
+
+	return 0;
+}
+
+static int mt8195_scp_before_load(struct mtk_scp *scp)
+{
+	/* clear SPM interrupt, SCP2SPM_IPC_CLR */
+	writel(0xff, scp->reg_base + MT8192_SCP2SPM_IPC_CLR);
+
+	writel(1, scp->reg_base + MT8192_CORE0_SW_RSTN_SET);
+
+	/* enable SRAM clock */
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			  MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);
+	scp_sram_power_on(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
 
 	/* enable MPU for all memory regions */
 	writel(0xff, scp->reg_base + MT8192_CORE0_MEM_ATT_PREDEF);
@@ -551,11 +572,25 @@ static void mt8183_scp_stop(struct mtk_scp *scp)
 static void mt8192_scp_stop(struct mtk_scp *scp)
 {
 	/* Disable SRAM clock */
-	mt8192_power_off_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_0);
-	mt8192_power_off_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_1);
-	mt8192_power_off_sram(scp->reg_base + MT8192_L2TCM_SRAM_PD_2);
-	mt8192_power_off_sram(scp->reg_base + MT8192_L1TCM_SRAM_PDN);
-	mt8192_power_off_sram(scp->reg_base + MT8192_CPU0_SRAM_PD);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L1TCM_SRAM_PDN, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
+
+	/* Disable SCP watchdog */
+	writel(0, scp->reg_base + MT8192_CORE0_WDT_CFG);
+}
+
+static void mt8195_scp_stop(struct mtk_scp *scp)
+{
+	/* Disable SRAM clock */
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			   MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);
+	scp_sram_power_off(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
 
 	/* Disable SCP watchdog */
 	writel(0, scp->reg_base + MT8192_CORE0_WDT_CFG);
@@ -901,11 +936,11 @@ static const struct mtk_scp_of_data mt8192_of_data = {
 
 static const struct mtk_scp_of_data mt8195_of_data = {
 	.scp_clk_get = mt8195_scp_clk_get,
-	.scp_before_load = mt8192_scp_before_load,
+	.scp_before_load = mt8195_scp_before_load,
 	.scp_irq_handler = mt8192_scp_irq_handler,
 	.scp_reset_assert = mt8192_scp_reset_assert,
 	.scp_reset_deassert = mt8192_scp_reset_deassert,
-	.scp_stop = mt8192_scp_stop,
+	.scp_stop = mt8195_scp_stop,
 	.scp_da_to_va = mt8192_scp_da_to_va,
 	.host_to_scp_reg = MT8192_GIPC_IN_SET,
 	.host_to_scp_int_bit = MT8192_HOST_IPC_INT_BIT,
Signature · 1 line
2.15.GIT
#2 lore raw

Il 16/03/22 17:34, Mathieu Poirier ha scritto:

Quoted 58 lines

Good morning,

Quoted 55 lines · Tinghan Shen

The definition of L1TCM_SRAM_PDN bits on mt8195 is different to mt8192.

L1TCM_SRAM_PDN bits[3:0] control the power of mt8195 L1TCM SRAM.

L1TCM_SRAM_PDN bits[7:4] control the access path to EMI for SCP. These bits have to be powered on to allow EMI access for SCP.

Bits[7:4] also affect audio DSP because audio DSP and SCP are placed on the same hardware bus. If SCP cannot access EMI, audio DSP is blocked too.

L1TCM_SRAM_PDN bits[31:8] are not used.

This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM. It's because the modification introduces a short period of time blocking audio DSP to access EMI. This was not a problem until we have to load both SCP module and audio DSP module. audio DSP needs to access EMI because it has source/data on DRAM. Audio DSP will have unexpected behavior when it accesses EMI and the SCP driver blocks the EMI path at the same time.

Fixes 79111df414fc ("remoteproc: mediatek: Support mt8195 scp")

---
v4: add Fixes and Reviewed-by tags
v3: fix build error
v2: apply comments about macro definition and function calls

---
  drivers/remoteproc/mtk_common.h |  2 ++
  drivers/remoteproc/mtk_scp.c    | 67 +++++++++++++++++++++++++++++++----------
  2 files changed, 53 insertions(+), 16 deletions(-)
drivers/remoteproc/mtk_common.h +2 -0
index 5ff3867c72f3..ff954a06637c 100644
@@ -51,6 +51,8 @@
  #define MT8192_CORE0_WDT_IRQ		0x10030
  #define MT8192_CORE0_WDT_CFG		0x10034
 
+#define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS		GENMASK(7, 4)
+
  #define SCP_FW_VER_LEN			32
  #define SCP_SHARE_BUFFER_SIZE		288
 
drivers/remoteproc/mtk_scp.c +0 -0
index 36e48cf58ed6..5f686fe09203 100644
@@ -365,22 +365,22 @@ static int mt8183_scp_before_load(struct mtk_scp *scp)
  	return 0;
  }
 
Quoted 7 lines

drivers/remoteproc/mtk_scp.c static int mt8183_scp_before_load(struct mtk_scp *scp) · line 368

Quoted 2 lines
drivers/remoteproc/mtk_scp.c +1 -1
-static void mt8192_power_on_sram(void __iomem *addr)
+static void scp_sram_power_on(void __iomem *addr, u32 reserved_mask)

Why is @reserved_mask needed? It is not described in the changelong and as far as I can see in this patchset the parameter is always set to '0', which has no effect on the mask that gets generated.

Hello Mathieu,
the @reserved_mask is explained in perhaps not very very clear terms, meaning that he's not explicitly saying the name of the new param, but that's it:

"This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM."

....and it's actually being used, check below....

Quoted 14 lines

Thanks,
Mathieu

Quoted 11 lines
drivers/remoteproc/mtk_scp.c +2 -2
  {
  	int i;
 
  	for (i = 31; i >= 0; i--)
-		writel(GENMASK(i, 0), addr);
+		writel(GENMASK(i, 0) & ~reserved_mask, addr);
  	writel(0, addr);
  }
 
-static void mt8192_power_off_sram(void __iomem *addr)
+static void scp_sram_power_off(void __iomem *addr, u32 reserved_mask)

...snip...

Quoted 13 lines
Quoted 11 lines
drivers/remoteproc/mtk_scp.c +11 -0
+static int mt8195_scp_before_load(struct mtk_scp *scp)
+{
+	/* clear SPM interrupt, SCP2SPM_IPC_CLR */
+	writel(0xff, scp->reg_base + MT8192_SCP2SPM_IPC_CLR);
+
+	writel(1, scp->reg_base + MT8192_CORE0_SW_RSTN_SET);
+
+	/* enable SRAM clock */
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
Quoted 2 lines

drivers/remoteproc/mtk_scp.c static int mt8192_scp_before_load(struct mtk_scp *scp) · line 418

Quoted 2 lines
drivers/remoteproc/mtk_scp.c +2 -0
+	scp_sram_power_on(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			  MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

here ^^^^^^^^^^^^

Quoted 4 lines
Quoted 4 lines
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_on(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
 
  	/* enable MPU for all memory regions */
  	writel(0xff, scp->reg_base + MT8192_CORE0_MEM_ATT_PREDEF);

...snip...

Quoted 9 lines

drivers/remoteproc/mtk_scp.c static void mt8183_scp_stop(struct mtk_scp *scp) · line 592

Quoted 9 lines
drivers/remoteproc/mtk_scp.c +9 -0
+
+static void mt8195_scp_stop(struct mtk_scp *scp)
+{
+	/* Disable SRAM clock */
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			   MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

and here ^^^^^^^^

Quoted 2 lines

drivers/remoteproc/mtk_scp.c

Quoted 2 lines
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_off(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);

Cheers,
Angelo

maintainer
#3 lore raw
Quoted 101 lines · AngeloGioacchino Del Regno

Il 16/03/22 17:34, Mathieu Poirier ha scritto:

Quoted 53 lines

Good morning,

Quoted 50 lines · Tinghan Shen

The definition of L1TCM_SRAM_PDN bits on mt8195 is different to mt8192.

L1TCM_SRAM_PDN bits[3:0] control the power of mt8195 L1TCM SRAM.

L1TCM_SRAM_PDN bits[7:4] control the access path to EMI for SCP. These bits have to be powered on to allow EMI access for SCP.

Bits[7:4] also affect audio DSP because audio DSP and SCP are placed on the same hardware bus. If SCP cannot access EMI, audio DSP is blocked too.

L1TCM_SRAM_PDN bits[31:8] are not used.

This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM. It's because the modification introduces a short period of time blocking audio DSP to access EMI. This was not a problem until we have to load both SCP module and audio DSP module. audio DSP needs to access EMI because it has source/data on DRAM. Audio DSP will have unexpected behavior when it accesses EMI and the SCP driver blocks the EMI path at the same time.

Fixes 79111df414fc ("remoteproc: mediatek: Support mt8195 scp")

---
v4: add Fixes and Reviewed-by tags
v3: fix build error
v2: apply comments about macro definition and function calls

---
  drivers/remoteproc/mtk_common.h |  2 ++
  drivers/remoteproc/mtk_scp.c    | 67 +++++++++++++++++++++++++++++++----------
  2 files changed, 53 insertions(+), 16 deletions(-)
drivers/remoteproc/mtk_common.h +2 -0
index 5ff3867c72f3..ff954a06637c 100644
@@ -51,6 +51,8 @@
  #define MT8192_CORE0_WDT_IRQ		0x10030
  #define MT8192_CORE0_WDT_CFG		0x10034
+#define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS		GENMASK(7, 4)
+
  #define SCP_FW_VER_LEN			32
  #define SCP_SHARE_BUFFER_SIZE		288
drivers/remoteproc/mtk_scp.c +0 -0
index 36e48cf58ed6..5f686fe09203 100644
@@ -365,22 +365,22 @@ static int mt8183_scp_before_load(struct mtk_scp *scp)
Quoted 9 lines

drivers/remoteproc/mtk_scp.c static int mt8183_scp_before_load(struct mtk_scp *scp) · line 368

Quoted 4 lines
drivers/remoteproc/mtk_scp.c +1 -1
  	return 0;
  }
-static void mt8192_power_on_sram(void __iomem *addr)
+static void scp_sram_power_on(void __iomem *addr, u32 reserved_mask)

Why is @reserved_mask needed? It is not described in the changelong and as far as I can see in this patchset the parameter is always set to '0', which has no effect on the mask that gets generated.

Hello Mathieu,
the @reserved_mask is explained in perhaps not very very clear terms, meaning that he's not explicitly saying the name of the new param, but that's it:

"This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM."

....and it's actually being used, check below....

Quoted 12 lines

Thanks,
Mathieu

Quoted 9 lines
drivers/remoteproc/mtk_scp.c +2 -2
  {
  	int i;
  	for (i = 31; i >= 0; i--)
-		writel(GENMASK(i, 0), addr);
+		writel(GENMASK(i, 0) & ~reserved_mask, addr);
  	writel(0, addr);
  }
-static void mt8192_power_off_sram(void __iomem *addr)
+static void scp_sram_power_off(void __iomem *addr, u32 reserved_mask)

...snip...

Quoted 11 lines
Quoted 11 lines
drivers/remoteproc/mtk_scp.c +11 -0
+static int mt8195_scp_before_load(struct mtk_scp *scp)
+{
+	/* clear SPM interrupt, SCP2SPM_IPC_CLR */
+	writel(0xff, scp->reg_base + MT8192_SCP2SPM_IPC_CLR);
+
+	writel(1, scp->reg_base + MT8192_CORE0_SW_RSTN_SET);
+
+	/* enable SRAM clock */
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
Quoted 4 lines
Quoted 2 lines

drivers/remoteproc/mtk_scp.c static int mt8192_scp_before_load(struct mtk_scp *scp) · line 418

Quoted 2 lines
drivers/remoteproc/mtk_scp.c +2 -0
+	scp_sram_power_on(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			  MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

here

Yes - it's obvious now that you point it out.

This patch conflicts with the newly added support for mt8186[1]. I tried to fix it but did not know if mt8185 needed the same kind of bit masking as mt8195. Please have a look and rebase to rproc-next.

Thanks,
Mathieu

[1]. 80d691854ffb remoteproc: mediatek: Support mt8186 scp

Quoted 23 lines
Quoted 4 lines
Quoted 4 lines
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_on(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
  	/* enable MPU for all memory regions */
  	writel(0xff, scp->reg_base + MT8192_CORE0_MEM_ATT_PREDEF);

...snip...

Quoted 9 lines

drivers/remoteproc/mtk_scp.c static void mt8183_scp_stop(struct mtk_scp *scp) · line 592

Quoted 9 lines
drivers/remoteproc/mtk_scp.c +9 -0
+
+static void mt8195_scp_stop(struct mtk_scp *scp)
+{
+	/* Disable SRAM clock */
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			   MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

and here ^^^^^^^^

Quoted 1 line

drivers/remoteproc/mtk_scp.c

Quoted 1 line
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_off(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);

Cheers,
Angelo

#4 lore raw

Hi Mathieu,

Quoted 107 lines · Mathieu Poirier
Quoted 101 lines · AngeloGioacchino Del Regno

Il 16/03/22 17:34, Mathieu Poirier ha scritto:

Quoted 53 lines

Good morning,

Quoted 50 lines · Tinghan Shen

The definition of L1TCM_SRAM_PDN bits on mt8195 is different to mt8192.

L1TCM_SRAM_PDN bits[3:0] control the power of mt8195 L1TCM SRAM.

L1TCM_SRAM_PDN bits[7:4] control the access path to EMI for SCP. These bits have to be powered on to allow EMI access for SCP.

Bits[7:4] also affect audio DSP because audio DSP and SCP are placed on the same hardware bus. If SCP cannot access EMI, audio DSP is blocked too.

L1TCM_SRAM_PDN bits[31:8] are not used.

This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM. It's because the modification introduces a short period of time blocking audio DSP to access EMI. This was not a problem until we have to load both SCP module and audio DSP module. audio DSP needs to access EMI because it has source/data on DRAM. Audio DSP will have unexpected behavior when it accesses EMI and the SCP driver blocks the EMI path at the same time.

Fixes 79111df414fc ("remoteproc: mediatek: Support mt8195 scp")

---
v4: add Fixes and Reviewed-by tags
v3: fix build error
v2: apply comments about macro definition and function calls

---
  drivers/remoteproc/mtk_common.h |  2 ++
  drivers/remoteproc/mtk_scp.c    | 67 +++++++++++++++++++++++++++++++----------
  2 files changed, 53 insertions(+), 16 deletions(-)
drivers/remoteproc/mtk_common.h +2 -0
index 5ff3867c72f3..ff954a06637c 100644
@@ -51,6 +51,8 @@
  #define MT8192_CORE0_WDT_IRQ		0x10030
  #define MT8192_CORE0_WDT_CFG		0x10034
+#define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS		GENMASK(7, 4)
+
  #define SCP_FW_VER_LEN			32
  #define SCP_SHARE_BUFFER_SIZE		288
drivers/remoteproc/mtk_scp.c +0 -0
index 36e48cf58ed6..5f686fe09203 100644
@@ -365,22 +365,22 @@ static int mt8183_scp_before_load(struct mtk_scp *scp)
Quoted 9 lines

drivers/remoteproc/mtk_scp.c static int mt8183_scp_before_load(struct mtk_scp *scp) · line 368

Quoted 4 lines
drivers/remoteproc/mtk_scp.c +1 -1
  	return 0;
  }
-static void mt8192_power_on_sram(void __iomem *addr)
+static void scp_sram_power_on(void __iomem *addr, u32 reserved_mask)

Why is @reserved_mask needed? It is not described in the changelong and as far as I can see in this patchset the parameter is always set to '0', which has no effect on the mask that gets generated.

Hello Mathieu,
the @reserved_mask is explained in perhaps not very very clear terms, meaning that he's not explicitly saying the name of the new param, but that's it:

"This fix removes modification of bits[7:4] when power on/off mt8195 SCP L1TCM."

....and it's actually being used, check below....

Quoted 12 lines

Thanks,
Mathieu

Quoted 9 lines
drivers/remoteproc/mtk_scp.c +2 -2
  {
  	int i;
  	for (i = 31; i >= 0; i--)
-		writel(GENMASK(i, 0), addr);
+		writel(GENMASK(i, 0) & ~reserved_mask, addr);
  	writel(0, addr);
  }
-static void mt8192_power_off_sram(void __iomem *addr)
+static void scp_sram_power_off(void __iomem *addr, u32 reserved_mask)

...snip...

Quoted 11 lines
Quoted 11 lines
drivers/remoteproc/mtk_scp.c +11 -0
+static int mt8195_scp_before_load(struct mtk_scp *scp)
+{
+	/* clear SPM interrupt, SCP2SPM_IPC_CLR */
+	writel(0xff, scp->reg_base + MT8192_SCP2SPM_IPC_CLR);
+
+	writel(1, scp->reg_base + MT8192_CORE0_SW_RSTN_SET);
+
+	/* enable SRAM clock */
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_on(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
Quoted 4 lines
Quoted 2 lines

drivers/remoteproc/mtk_scp.c static int mt8192_scp_before_load(struct mtk_scp *scp) · line 418

Quoted 2 lines
drivers/remoteproc/mtk_scp.c +2 -0
+	scp_sram_power_on(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			  MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

here

Quoted 8 lines

Yes - it's obvious now that you point it out.

This patch conflicts with the newly added support for mt8186[1]. I tried to fix it but did not know if mt8185 needed the same kind of bit masking as mt8195. Please have a look and rebase to rproc-next.

Thanks,
Mathieu

Ok, I'll rebase to rproc-next at next version.
Thank you!

Best regards,
TingHan

Quoted 26 lines

[1]. 80d691854ffb remoteproc: mediatek: Support mt8186 scp

Quoted 23 lines
Quoted 4 lines
Quoted 4 lines
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_on(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);
  	/* enable MPU for all memory regions */
  	writel(0xff, scp->reg_base + MT8192_CORE0_MEM_ATT_PREDEF);

...snip...

Quoted 9 lines

drivers/remoteproc/mtk_scp.c static void mt8183_scp_stop(struct mtk_scp *scp) · line 592

Quoted 9 lines
drivers/remoteproc/mtk_scp.c +9 -0
+
+static void mt8195_scp_stop(struct mtk_scp *scp)
+{
+	/* Disable SRAM clock */
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_0, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_1, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L2TCM_SRAM_PD_2, 0);
+	scp_sram_power_off(scp->reg_base + MT8192_L1TCM_SRAM_PDN,
+			   MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS);

and here ^^^^^^^^

Quoted 1 line

drivers/remoteproc/mtk_scp.c

Quoted 1 line
drivers/remoteproc/mtk_scp.c +1 -0
+	scp_sram_power_off(scp->reg_base + MT8192_CPU0_SRAM_PD, 0);

Cheers,
Angelo