Skip to contentSkip to conversation

i2c: piix4: Replace cd6h/cd7h port I/O accesses with MMIO accesses

PATCH Superseded Terry Bowman 4 participants 27 messages Jan 30, 2022 – Feb 9, 2022

Subsystems (3)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

v4 cover i2c: piix4: Replace cd6h/cd7h port I/O accesses with MMIO accesses

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

This series includes patches with MMIO accesses to register FCH::PM::DECODEEN. The same register is also accessed by the sp5100_tco driver.[1] Synchronization to the MMIO register is required in both drivers.

The first patch creates a macro to request MMIO region using the 'muxed' retry logic. This is used in patch 6 to synchronize accesses to EFCH MMIO.

The second patch replaces a hardcoded region size with a #define. This is to improve maintainability and was requested from v2 review.

The third patch moves duplicated region request/release code into functions. This locates related code into functions and reduces code line count. This will also make adding MMIO support in patch 6 easier.

The fourth patch moves SMBus controller address detection into a function. This is in preparation for adding MMIO region support.

The fifth patch moves EFCH port selection into a function. This is in preparation for adding MMIO region support.

The sixth patch adds MMIO support for region requesting/releasing and mapping. This is necessary for using MMIO to detect SMBus controller address, enable SMBbus controller region, and control the port select.

The seventh patch updates the SMBus controller address detection to support using MMIO. This is necessary because the driver accesses register FCH::PM::DECODEEN during initialization and only available using MMIO on later AMD processors.

The eighth patch updates the SMBus port selection to support MMIO. This is required because port selection control resides in the FCH::PM::DECODEEN[smbus0sel] and is only accessible using MMIO on later AMD processors.

The ninth patch enables the EFCH MMIO functionality added earlier in this series. The SMBus controller's PCI revision ID is used to check if EFCH MMIO is supported by HW and should be enabled in the driver.

Based on v5.16.

Testing:
Tested on family 19h using:
i2cdetect -y 0
i2cdetect -y 2

- Results using v5.16 and this pachset applied. Below

    shows the devices detected on the busses:
    
    # i2cdetect -y 0 
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- -- 
    10: 10 11 -- -- -- -- -- -- 18 -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: 30 -- -- -- -- 35 36 -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- -- 
    50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    70: -- -- -- 73 -- -- -- --                         
    # i2cdetect -y 2
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- -- 
    10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- -- 
    50: -- 51 -- -- 54 -- -- -- -- -- -- -- -- -- -- -- 
    60: 60 -- -- 63 -- -- 66 -- -- -- -- 6b -- -- 6e -- 
    70: 70 71 72 73 74 75 -- 77

Also tested using sp5100_tco submitted series listed below.[1] I applied the sp5100_tco v4 series and ran:
cat >> /dev/watchdog

[1] sp5100_tco v4 patchset is in process but not sent yet. Below is v3 upstream review:

Changes in v4:
- Changed request_muxed_mem_region() macro to request_mem_region_muxed() in patch 1. (Andy Shevchenko)
- Removed unnecessary newline where possible in calls to request_muxed_region() in patch 2. (Terry Bowman)
- Changed piix4_sb800_region_setup() to piix4_sb800_region_request(). Patch 3. (Jean Delvare)
- Reordered piix4_setup_sb800() local variables from longest name length to shortest name length. Patch 4. (Terry Bowman)
- Changed piix4_sb800_region_request() and piix4_sb800_region_release() by adding early return() to remove 'else' improving readability. Patch 6. (Terry Bowman)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)
- Refactored piix4_sb800_port_sel() to simplify the 'if' statement using temp variable. Patch 8. (Terry Bowman)
- Added mmio_cfg.use_mmio assignment in piix4_add_adapter(). This is needed for calls to piix4_sb800_port_sel() after initialization during normal operation. Patch 9. (Terry Bowman)

Changes in v3:
- Added request_muxed_mem_region() patch (Wolfram, Guenter)
- Reduced To/Cc list length. (Andy)

Changes in v2:
- Split single patch. (Jean Delvare)
- Replace constant 2 with SB800_PIIX4_SMB_MAP_SIZE where appropriate. (Jean Delvare)
- Shorten SB800_PIIX4_FCH_PM_DECODEEN_MMIO_EN name length to SB800_PIIX4_FCH_PM_DECODEEN_MMIO. (Jean Delvare)
- Change AMD_PCI_SMBUS_REVISION_MMIO from 0x59 to 0x51. (Terry Bowman)
- Change piix4_sb800_region_setup() to piix4_sb800_region_request(). (Jean Delvare)
- Change 'SMB' text in logging to 'SMBus' (Jean Delvare)
- Remove unnecessary NULL assignment in piix4_sb800_region_release(). (Jean Delvare)
- Move 'u8' variable definitions to single line. (Jean Delvare)
- Hardcode piix4_setup_sb800_smba() return value to 0 since it is always
0. (Jean Delvare)

Terry Bowman (9):
kernel/resource: Introduce request_mem_region_muxed()
i2c: piix4: Replace hardcoded memory map size with a #define
i2c: piix4: Move port I/O region request/release code into functions
i2c: piix4: Move SMBus controller base address detect into function
i2c: piix4: Move SMBus port selection into function
i2c: piix4: Add EFCH MMIO support to region request and release
i2c: piix4: Add EFCH MMIO support to SMBus base address detect
i2c: piix4: Add EFCH MMIO support for SMBus port select
i2c: piix4: Enable EFCH MMIO for Family 17h+

 drivers/i2c/busses/i2c-piix4.c | 207 ++++++++++++++++++++++++++-------
 include/linux/ioport.h         |   2 +
 2 files changed, 164 insertions(+), 45 deletions(-)
Signature · 1 line
2.30.2
#1 lore raw

PATCH v4 1/9 kernel/resource: Introduce request_mem_region_muxed()

Support for requesting muxed memory region is implemented but not currently callable as a macro. Add the request muxed memory region macro.

MMIO memory accesses can be synchronized using request_mem_region() which is already available. This call will return failure if the resource is busy. The 'muxed' version of this macro will handle a busy resource by using a wait queue to retry until the resource is available.

---
 include/linux/ioport.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/ioport.h +2 -0
index 8359c50f9988..ec5f71f7135b 100644
@@ -262,6 +262,8 @@ resource_union(struct resource *r1, struct resource *r2, struct resource *r)
 #define request_muxed_region(start,n,name)	__request_region(&ioport_resource, (start), (n), (name), IORESOURCE_MUXED)
 #define __request_mem_region(start,n,name, excl) __request_region(&iomem_resource, (start), (n), (name), excl)
 #define request_mem_region(start,n,name) __request_region(&iomem_resource, (start), (n), (name), 0)
+#define request_mem_region_muxed(start, n, name) \
+	__request_region(&iomem_resource, (start), (n), (name), IORESOURCE_MUXED)
 #define request_mem_region_exclusive(start,n,name) \
 	__request_region(&iomem_resource, (start), (n), (name), IORESOURCE_EXCLUSIVE)
 #define rename_region(region, newname) do { (region)->name = (newname); } while (0)
Signature · 1 line
2.30.2
#2 lore raw

PATCH v4 2/9 i2c: piix4: Replace hardcoded memory map size with a #define

In reply to Terry Bowman

Replace number constant with #define to improve readability and maintainability.

---
 drivers/i2c/busses/i2c-piix4.c | 16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)
drivers/i2c/busses/i2c-piix4.c +10 -6
index 8c1b31ed0c42..3ff68967034e 100644
@@ -77,6 +77,7 @@
 
 /* SB800 constants */
 #define SB800_PIIX4_SMB_IDX		0xcd6
+#define SB800_PIIX4_SMB_MAP_SIZE	2
 
 #define KERNCZ_IMC_IDX			0x3e
 #define KERNCZ_IMC_DATA			0x3f
@@ -290,7 +291,8 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	else
 		smb_en = (aux) ? 0x28 : 0x2c;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, 2, "sb800_piix4_smb")) {
+	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
+				  "sb800_piix4_smb")) {
 		dev_err(&PIIX4_dev->dev,
 			"SMB base address index region 0x%x already in use.\n",
 			SB800_PIIX4_SMB_IDX);
@@ -302,7 +304,7 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
 	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
-	release_region(SB800_PIIX4_SMB_IDX, 2);
+	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 
 	if (!smb_en) {
 		smb_en_status = smba_en_lo & 0x10;
@@ -371,7 +373,8 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 			piix4_port_shift_sb800 = SB800_PIIX4_PORT_IDX_SHIFT;
 		}
 	} else {
-		if (!request_muxed_region(SB800_PIIX4_SMB_IDX, 2,
+		if (!request_muxed_region(SB800_PIIX4_SMB_IDX,
+					  SB800_PIIX4_SMB_MAP_SIZE,
 					  "sb800_piix4_smb")) {
 			release_region(piix4_smba, SMBIOSIZE);
 			return -EBUSY;
@@ -384,7 +387,7 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 				       SB800_PIIX4_PORT_IDX;
 		piix4_port_mask_sb800 = SB800_PIIX4_PORT_IDX_MASK;
 		piix4_port_shift_sb800 = SB800_PIIX4_PORT_IDX_SHIFT;
-		release_region(SB800_PIIX4_SMB_IDX, 2);
+		release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 	}
 
 	dev_info(&PIIX4_dev->dev,
@@ -682,7 +685,8 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	u8 port;
 	int retval;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, 2, "sb800_piix4_smb"))
+	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
+				  "sb800_piix4_smb"))
 		return -EBUSY;
 
 	/* Request the SMBUS semaphore, avoid conflicts with the IMC */
@@ -758,7 +762,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		piix4_imc_wakeup();
 
 release:
-	release_region(SB800_PIIX4_SMB_IDX, 2);
+	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 	return retval;
 }
 
Signature · 1 line
2.30.2
#3 lore raw

PATCH v4 3/9 i2c: piix4: Move port I/O region request/release code into functions

In reply to Terry Bowman

Move duplicated region request and release code into a function. Move is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 39 +++++++++++++++++++++++-----------
 1 file changed, 27 insertions(+), 12 deletions(-)
drivers/i2c/busses/i2c-piix4.c +27 -12
index 3ff68967034e..5a98970ac60a 100644
@@ -165,6 +165,24 @@ struct i2c_piix4_adapdata {
 	u8 port;		/* Port number, shifted */
 };
 
+static int piix4_sb800_region_request(struct device *dev)
+{
+	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
+				  "sb800_piix4_smb")) {
+		dev_err(dev,
+			"SMBus base address index region 0x%x already in use.\n",
+			SB800_PIIX4_SMB_IDX);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static void piix4_sb800_region_release(struct device *dev)
+{
+	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+}
+
 static int piix4_setup(struct pci_dev *PIIX4_dev,
 		       const struct pci_device_id *id)
 {
@@ -270,6 +288,7 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	unsigned short piix4_smba;
 	u8 smba_en_lo, smba_en_hi, smb_en, smb_en_status, port_sel;
 	u8 i2ccfg, i2ccfg_offset = 0x10;
+	int retval;
 
 	/* SB800 and later SMBus does not support forcing address */
 	if (force || force_addr) {
@@ -291,20 +310,16 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	else
 		smb_en = (aux) ? 0x28 : 0x2c;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb")) {
-		dev_err(&PIIX4_dev->dev,
-			"SMB base address index region 0x%x already in use.\n",
-			SB800_PIIX4_SMB_IDX);
-		return -EBUSY;
-	}
+	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	if (retval)
+		return retval;
 
 	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
 	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
 	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
 	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&PIIX4_dev->dev);
 
 	if (!smb_en) {
 		smb_en_status = smba_en_lo & 0x10;
@@ -685,9 +700,9 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	u8 port;
 	int retval;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb"))
-		return -EBUSY;
+	retval = piix4_sb800_region_request(&adap->dev);
+	if (retval)
+		return retval;
 
 	/* Request the SMBUS semaphore, avoid conflicts with the IMC */
 	smbslvcnt  = inb_p(SMBSLVCNT);
@@ -762,7 +777,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		piix4_imc_wakeup();
 
 release:
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&adap->dev);
 	return retval;
 }
 
Signature · 1 line
2.30.2
#4 lore raw

Re: [PATCH v4 3/9] i2c: piix4: Move port I/O region request/release code into functions

Hi Terry,

Quoted 84 lines · Terry Bowman

Move duplicated region request and release code into a function. Move is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 39 +++++++++++++++++++++++-----------
 1 file changed, 27 insertions(+), 12 deletions(-)
drivers/i2c/busses/i2c-piix4.c +26 -11
index 3ff68967034e..5a98970ac60a 100644
@@ -165,6 +165,24 @@ struct i2c_piix4_adapdata {
 	u8 port;		/* Port number, shifted */
 };
 
+static int piix4_sb800_region_request(struct device *dev)
+{
+	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
+				  "sb800_piix4_smb")) {
+		dev_err(dev,
+			"SMBus base address index region 0x%x already in use.\n",
+			SB800_PIIX4_SMB_IDX);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static void piix4_sb800_region_release(struct device *dev)
+{
+	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+}
+
 static int piix4_setup(struct pci_dev *PIIX4_dev,
 		       const struct pci_device_id *id)
 {
@@ -270,6 +288,7 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	unsigned short piix4_smba;
 	u8 smba_en_lo, smba_en_hi, smb_en, smb_en_status, port_sel;
 	u8 i2ccfg, i2ccfg_offset = 0x10;
+	int retval;
 
 	/* SB800 and later SMBus does not support forcing address */
 	if (force || force_addr) {
@@ -291,20 +310,16 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	else
 		smb_en = (aux) ? 0x28 : 0x2c;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb")) {
-		dev_err(&PIIX4_dev->dev,
-			"SMB base address index region 0x%x already in use.\n",
-			SB800_PIIX4_SMB_IDX);
-		return -EBUSY;
-	}
+	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	if (retval)
+		return retval;
 
 	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
 	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
 	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
 	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&PIIX4_dev->dev);
 
 	if (!smb_en) {
 		smb_en_status = smba_en_lo & 0x10;
@@ -685,9 +700,9 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	u8 port;
 	int retval;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb"))
-		return -EBUSY;
+	retval = piix4_sb800_region_request(&adap->dev);
+	if (retval)
+		return retval;
 
 	/* Request the SMBUS semaphore, avoid conflicts with the IMC */
 	smbslvcnt  = inb_p(SMBSLVCNT);

drivers/i2c/busses/i2c-piix4.c static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr, · line 783

Quoted 8 lines
drivers/i2c/busses/i2c-piix4.c +1 -1
@@ -762,7 +777,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		piix4_imc_wakeup();
 
 release:
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&adap->dev);
 	return retval;
 }
 

There's a third occurrence of request_muxed_region(SB800_PIIX4_SMB_IDX, ...) / release_region(SB800_PIIX4_SMB_IDX, ...) in function piix4_setup_sb800. Any reason why you don't make use of the new helper functions there as well?

OK, I see that this part of the code is specific to the original (ATI) SB800, so it can't use MMIO, therefore you don't *have* to call the helper functions. But for consistency, wouldn't it still make sense to use them?

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#5 lore raw

Re: [PATCH v4 3/9] i2c: piix4: Move port I/O region request/release code into functions

Hi Jean,

Quoted 89 lines · Jean Delvare

Hi Terry,

Quoted 86 lines · Terry Bowman

Move duplicated region request and release code into a function. Move is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 39 +++++++++++++++++++++++-----------
 1 file changed, 27 insertions(+), 12 deletions(-)
drivers/i2c/busses/i2c-piix4.c +26 -11
index 3ff68967034e..5a98970ac60a 100644
@@ -165,6 +165,24 @@ struct i2c_piix4_adapdata {
 	u8 port;		/* Port number, shifted */
 };
 
+static int piix4_sb800_region_request(struct device *dev)
+{
+	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
+				  "sb800_piix4_smb")) {
+		dev_err(dev,
+			"SMBus base address index region 0x%x already in use.\n",
+			SB800_PIIX4_SMB_IDX);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static void piix4_sb800_region_release(struct device *dev)
+{
+	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+}
+
 static int piix4_setup(struct pci_dev *PIIX4_dev,
 		       const struct pci_device_id *id)
 {
@@ -270,6 +288,7 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	unsigned short piix4_smba;
 	u8 smba_en_lo, smba_en_hi, smb_en, smb_en_status, port_sel;
 	u8 i2ccfg, i2ccfg_offset = 0x10;
+	int retval;
 
 	/* SB800 and later SMBus does not support forcing address */
 	if (force || force_addr) {
@@ -291,20 +310,16 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	else
 		smb_en = (aux) ? 0x28 : 0x2c;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb")) {
-		dev_err(&PIIX4_dev->dev,
-			"SMB base address index region 0x%x already in use.\n",
-			SB800_PIIX4_SMB_IDX);
-		return -EBUSY;
-	}
+	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	if (retval)
+		return retval;
 
 	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
 	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
 	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
 	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&PIIX4_dev->dev);
 
 	if (!smb_en) {
 		smb_en_status = smba_en_lo & 0x10;
@@ -685,9 +700,9 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	u8 port;
 	int retval;
 
-	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
-				  "sb800_piix4_smb"))
-		return -EBUSY;
+	retval = piix4_sb800_region_request(&adap->dev);
+	if (retval)
+		return retval;
 
 	/* Request the SMBUS semaphore, avoid conflicts with the IMC */
 	smbslvcnt  = inb_p(SMBSLVCNT);
@@ -762,7 +777,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		piix4_imc_wakeup();
 
Quoted 12 lines

drivers/i2c/busses/i2c-piix4.c static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr, · line 783

Quoted 6 lines
drivers/i2c/busses/i2c-piix4.c +1 -1
 release:
-	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
+	piix4_sb800_region_release(&adap->dev);
 	return retval;
 }
 

There's a third occurrence of request_muxed_region(SB800_PIIX4_SMB_IDX, ...) / release_region(SB800_PIIX4_SMB_IDX, ...) in function piix4_setup_sb800. Any reason why you don't make use of the new helper functions there as well?

I didn't update the other occurrence because it was outside the codepath for the device we are addressing. At the time I wanted to minimize changes particularly for other devices.

Quoted 5 lines

OK, I see that this part of the code is specific to the original (ATI) SB800, so it can't use MMIO, therefore you don't *have* to call the helper functions. But for consistency, wouldn't it still make sense to use them?

Yes, it would be more consistent if it used the helper function.

Regards,
Terry

#6 lore raw

PATCH v4 4/9 i2c: piix4: Move SMBus controller base address detect into function

In reply to Terry Bowman

Move SMBus controller base address detection into function. Refactor is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 71 +++++++++++++++++++++-------------
 1 file changed, 45 insertions(+), 26 deletions(-)
drivers/i2c/busses/i2c-piix4.c +45 -26
index 5a98970ac60a..eab06e9e5fdf 100644
@@ -282,12 +282,52 @@ static int piix4_setup(struct pci_dev *PIIX4_dev,
 	return piix4_smba;
 }
 
+static int piix4_setup_sb800_smba(struct pci_dev *PIIX4_dev,
+				  u8 smb_en,
+				  u8 aux,
+				  u8 *smb_en_status,
+				  unsigned short *piix4_smba)
+{
+	u8 smba_en_lo;
+	u8 smba_en_hi;
+	int retval;
+
+	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	if (retval)
+		return retval;
+
+	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
+	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
+	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
+	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
+
+	piix4_sb800_region_release(&PIIX4_dev->dev);
+
+	if (!smb_en) {
+		*smb_en_status = smba_en_lo & 0x10;
+		*piix4_smba = smba_en_hi << 8;
+		if (aux)
+			*piix4_smba |= 0x20;
+	} else {
+		*smb_en_status = smba_en_lo & 0x01;
+		*piix4_smba = ((smba_en_hi << 8) | smba_en_lo) & 0xffe0;
+	}
+
+	if (!*smb_en_status) {
+		dev_err(&PIIX4_dev->dev,
+			"SMBus Host Controller not enabled!\n");
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
 static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 			     const struct pci_device_id *id, u8 aux)
 {
-	unsigned short piix4_smba;
-	u8 smba_en_lo, smba_en_hi, smb_en, smb_en_status, port_sel;
+	u8 smb_en, smb_en_status, port_sel;
 	u8 i2ccfg, i2ccfg_offset = 0x10;
+	unsigned short piix4_smba;
 	int retval;
 
 	/* SB800 and later SMBus does not support forcing address */
@@ -310,33 +350,12 @@ static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,
 	else
 		smb_en = (aux) ? 0x28 : 0x2c;
 
-	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	retval = piix4_setup_sb800_smba(PIIX4_dev, smb_en, aux, &smb_en_status,
+					&piix4_smba);
+
 	if (retval)
 		return retval;
 
-	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
-	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
-	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
-	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
-
-	piix4_sb800_region_release(&PIIX4_dev->dev);
-
-	if (!smb_en) {
-		smb_en_status = smba_en_lo & 0x10;
-		piix4_smba = smba_en_hi << 8;
-		if (aux)
-			piix4_smba |= 0x20;
-	} else {
-		smb_en_status = smba_en_lo & 0x01;
-		piix4_smba = ((smba_en_hi << 8) | smba_en_lo) & 0xffe0;
-	}
-
-	if (!smb_en_status) {
-		dev_err(&PIIX4_dev->dev,
-			"SMBus Host Controller not enabled!\n");
-		return -ENODEV;
-	}
-
 	if (acpi_check_region(piix4_smba, SMBIOSIZE, piix4_driver.name))
 		return -ENODEV;
 
Signature · 1 line
2.30.2
#7 lore raw

Re: [PATCH v4 4/9] i2c: piix4: Move SMBus controller base address detect into function

drivers/i2c/busses/i2c-piix4.c static int piix4_setup(struct pci_dev *PIIX4_dev, · line 333

Quoted 11 lines · Terry Bowman

static int piix4_setup_sb800(struct pci_dev *PIIX4_dev,

drivers/i2c/busses/i2c-piix4.c +2 -2
 			     const struct pci_device_id *id, u8 aux)
 {
-	unsigned short piix4_smba;
-	u8 smba_en_lo, smba_en_hi, smb_en, smb_en_status, port_sel;
+	u8 smb_en, smb_en_status, port_sel;
 	u8 i2ccfg, i2ccfg_offset = 0x10;
+	unsigned short piix4_smba;
 	int retval;
 
 	/* SB800 and later SMBus does not support forcing address */

I suggest not moving variables around unless there's a compelling to do that. Else you make your patches larger (and slightly harder to review), plus it increases the risks of context conflicts when backporting other changes.

Thanks,

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#8 lore raw

PATCH v4 5/9 i2c: piix4: Move SMBus port selection into function

In reply to Terry Bowman

Move port selection code into a separate function. Refactor is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 26 ++++++++++++++++----------
 1 file changed, 16 insertions(+), 10 deletions(-)
drivers/i2c/busses/i2c-piix4.c +16 -10
index eab06e9e5fdf..32a30af5778a 100644
@@ -699,6 +699,19 @@ static void piix4_imc_wakeup(void)
 	release_region(KERNCZ_IMC_IDX, 2);
 }
 
+static int piix4_sb800_port_sel(u8 port)
+{
+	u8 smba_en_lo, val;
+
+	outb_p(piix4_port_sel_sb800, SB800_PIIX4_SMB_IDX);
+	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
+
+	val = (smba_en_lo & ~piix4_port_mask_sb800) | port;
+	if (smba_en_lo != val)
+		outb_p(val, SB800_PIIX4_SMB_IDX + 1);
+
+	return (smba_en_lo & piix4_port_mask_sb800);
+}
 /*
  * Handles access to multiple SMBus ports on the SB800.
  * The port is selected by bits 2:1 of the smb_en register (0x2c).
@@ -715,8 +728,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	unsigned short piix4_smba = adapdata->smba;
 	int retries = MAX_TIMEOUT;
 	int smbslvcnt;
-	u8 smba_en_lo;
-	u8 port;
+	u8 prev_port;
 	int retval;
 
 	retval = piix4_sb800_region_request(&adap->dev);
@@ -776,18 +788,12 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		}
 	}
 
-	outb_p(piix4_port_sel_sb800, SB800_PIIX4_SMB_IDX);
-	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
-
-	port = adapdata->port;
-	if ((smba_en_lo & piix4_port_mask_sb800) != port)
-		outb_p((smba_en_lo & ~piix4_port_mask_sb800) | port,
-		       SB800_PIIX4_SMB_IDX + 1);
+	prev_port = piix4_sb800_port_sel(adapdata->port);
 
 	retval = piix4_access(adap, addr, flags, read_write,
 			      command, size, data);
 
-	outb_p(smba_en_lo, SB800_PIIX4_SMB_IDX + 1);
+	piix4_sb800_port_sel(prev_port);
 
 	/* Release the semaphore */
 	outb_p(smbslvcnt | 0x20, SMBSLVCNT);
Signature · 1 line
2.30.2
#9 lore raw

Re: [PATCH v4 5/9] i2c: piix4: Move SMBus port selection into function

Quoted 53 lines · Terry Bowman

Move port selection code into a separate function. Refactor is in preparation for following MMIO changes.

---
 drivers/i2c/busses/i2c-piix4.c | 26 ++++++++++++++++----------
 1 file changed, 16 insertions(+), 10 deletions(-)
drivers/i2c/busses/i2c-piix4.c +14 -9
index eab06e9e5fdf..32a30af5778a 100644
@@ -699,6 +699,19 @@ static void piix4_imc_wakeup(void)
 	release_region(KERNCZ_IMC_IDX, 2);
 }
 
+static int piix4_sb800_port_sel(u8 port)
+{
+	u8 smba_en_lo, val;
+
+	outb_p(piix4_port_sel_sb800, SB800_PIIX4_SMB_IDX);
+	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
+
+	val = (smba_en_lo & ~piix4_port_mask_sb800) | port;
+	if (smba_en_lo != val)
+		outb_p(val, SB800_PIIX4_SMB_IDX + 1);
+
+	return (smba_en_lo & piix4_port_mask_sb800);
+}
 /*
  * Handles access to multiple SMBus ports on the SB800.
  * The port is selected by bits 2:1 of the smb_en register (0x2c).
@@ -715,8 +728,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	unsigned short piix4_smba = adapdata->smba;
 	int retries = MAX_TIMEOUT;
 	int smbslvcnt;
-	u8 smba_en_lo;
-	u8 port;
+	u8 prev_port;
 	int retval;
 
 	retval = piix4_sb800_region_request(&adap->dev);
@@ -776,18 +788,12 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		}
 	}
 
-	outb_p(piix4_port_sel_sb800, SB800_PIIX4_SMB_IDX);
-	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
-
-	port = adapdata->port;
-	if ((smba_en_lo & piix4_port_mask_sb800) != port)
-		outb_p((smba_en_lo & ~piix4_port_mask_sb800) | port,
-		       SB800_PIIX4_SMB_IDX + 1);

drivers/i2c/busses/i2c-piix4.c static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr, · line 799

Quoted 10 lines
drivers/i2c/busses/i2c-piix4.c +2 -1
+	prev_port = piix4_sb800_port_sel(adapdata->port);
 
 	retval = piix4_access(adap, addr, flags, read_write,
 			      command, size, data);
 
-	outb_p(smba_en_lo, SB800_PIIX4_SMB_IDX + 1);
+	piix4_sb800_port_sel(prev_port);
 
 	/* Release the semaphore */
 	outb_p(smbslvcnt | 0x20, SMBSLVCNT);

Hmm, this gets pretty inefficient. You set SB800_PIIX4_SMB_IDX to piix4_port_sel_sb800 twice and compare the new port with the previous port twice, even though the result will inevitably be the same. The original code would just unconditionally restore the original port value, which was also not always needed, but was cheaper in the end.

I can't really think of a better way though, so, so be it :-(

I wonder why we insist on restoring the port though. When you read multiple values from the same bus (and I2C device drivers do that a lot, obviously) you end up switching ports a lot for no good reason. I looked at the git history and it has been that way since support for the multiplexed buses was added back in 2015. There's a mutex protecting that section of code anyway, so I don't think restoring the port buys us anything we did not already have. The only benefit I can see is to leave the port setting in its original state on suspend and shutdown (and we know by experience that it can be important to the BIOS) but then we should just do it in the suspend and remove functions, instead of after every transfer.

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#10 lore raw

PATCH v4 6/9 i2c: piix4: Add EFCH MMIO support to region request and release

In reply to Terry Bowman

EFCH cd6h/cd7h port I/O may no longer be available on later AMD processors and it is recommended to use MMIO instead. Update the request and release functions to support MMIO.

MMIO request/release and mmapping require details during cleanup. Add a MMIO configuration structure containing resource and vaddress details for mapping the region, accessing the region, and releasing the region.

---
 drivers/i2c/busses/i2c-piix4.c | 61 ++++++++++++++++++++++++++++++----
 1 file changed, 55 insertions(+), 6 deletions(-)
drivers/i2c/busses/i2c-piix4.c +55 -6
index 32a30af5778a..7defa0c5f1f9 100644
@@ -98,6 +98,9 @@
 #define SB800_PIIX4_PORT_IDX_MASK_KERNCZ	0x18
 #define SB800_PIIX4_PORT_IDX_SHIFT_KERNCZ	3
 
+#define SB800_PIIX4_FCH_PM_ADDR			0xFED80300
+#define SB800_PIIX4_FCH_PM_SIZE			8
+
 /* insmod parameters */
 
 /* If force is set to anything different from 0, we forcibly enable the
@@ -156,6 +159,12 @@ static const char *piix4_main_port_names_sb800[PIIX4_MAX_ADAPTERS] = {
 };
 static const char *piix4_aux_port_name_sb800 = " port 1";
 
+struct sb800_mmio_cfg {
+	void __iomem *addr;
+	struct resource *res;
+	bool use_mmio;
+};
+
 struct i2c_piix4_adapdata {
 	unsigned short smba;
 
@@ -163,10 +172,40 @@ struct i2c_piix4_adapdata {
 	bool sb800_main;
 	bool notify_imc;
 	u8 port;		/* Port number, shifted */
+	struct sb800_mmio_cfg mmio_cfg;
 };
 
-static int piix4_sb800_region_request(struct device *dev)
+static int piix4_sb800_region_request(struct device *dev,
+				      struct sb800_mmio_cfg *mmio_cfg)
 {
+	if (mmio_cfg->use_mmio) {
+		struct resource *res;
+		void __iomem *addr;
+
+		res = request_mem_region_muxed(SB800_PIIX4_FCH_PM_ADDR,
+					       SB800_PIIX4_FCH_PM_SIZE,
+					       "sb800_piix4_smb");
+		if (!res) {
+			dev_err(dev,
+				"SMBus base address memory region 0x%x already in use.\n",
+				SB800_PIIX4_FCH_PM_ADDR);
+			return -EBUSY;
+		}
+
+		addr = ioremap(SB800_PIIX4_FCH_PM_ADDR,
+			       SB800_PIIX4_FCH_PM_SIZE);
+		if (!addr) {
+			release_resource(res);
+			dev_err(dev, "SMBus base address mapping failed.\n");
+			return -ENOMEM;
+		}
+
+		mmio_cfg->res = res;
+		mmio_cfg->addr = addr;
+
+		return 0;
+	}
+
 	if (!request_muxed_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE,
 				  "sb800_piix4_smb")) {
 		dev_err(dev,
@@ -178,8 +217,15 @@ static int piix4_sb800_region_request(struct device *dev)
 	return 0;
 }
 
-static void piix4_sb800_region_release(struct device *dev)
+static void piix4_sb800_region_release(struct device *dev,
+				       struct sb800_mmio_cfg *mmio_cfg)
 {
+	if (mmio_cfg->use_mmio) {
+		iounmap(mmio_cfg->addr);
+		release_resource(mmio_cfg->res);
+		return;
+	}
+
 	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 }
 
@@ -288,11 +334,13 @@ static int piix4_setup_sb800_smba(struct pci_dev *PIIX4_dev,
 				  u8 *smb_en_status,
 				  unsigned short *piix4_smba)
 {
+	struct sb800_mmio_cfg mmio_cfg;
 	u8 smba_en_lo;
 	u8 smba_en_hi;
 	int retval;
 
-	retval = piix4_sb800_region_request(&PIIX4_dev->dev);
+	mmio_cfg.use_mmio = 0;
+	retval = piix4_sb800_region_request(&PIIX4_dev->dev, &mmio_cfg);
 	if (retval)
 		return retval;
 
@@ -301,7 +349,7 @@ static int piix4_setup_sb800_smba(struct pci_dev *PIIX4_dev,
 	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
 	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
-	piix4_sb800_region_release(&PIIX4_dev->dev);
+	piix4_sb800_region_release(&PIIX4_dev->dev, &mmio_cfg);
 
 	if (!smb_en) {
 		*smb_en_status = smba_en_lo & 0x10;
@@ -731,7 +779,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 	u8 prev_port;
 	int retval;
 
-	retval = piix4_sb800_region_request(&adap->dev);
+	retval = piix4_sb800_region_request(&adap->dev, &adapdata->mmio_cfg);
 	if (retval)
 		return retval;
 
@@ -802,7 +850,7 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		piix4_imc_wakeup();
 
 release:
-	piix4_sb800_region_release(&adap->dev);
+	piix4_sb800_region_release(&adap->dev, &adapdata->mmio_cfg);
 	return retval;
 }
 
@@ -880,6 +928,7 @@ static int piix4_add_adapter(struct pci_dev *dev, unsigned short smba,
 		return -ENOMEM;
 	}
 
+	adapdata->mmio_cfg.use_mmio = 0;
 	adapdata->smba = smba;
 	adapdata->sb800_main = sb800_main;
 	adapdata->port = port << piix4_port_shift_sb800;
Signature · 1 line
2.30.2
#11 lore raw

Re: [PATCH v4 6/9] i2c: piix4: Add EFCH MMIO support to region request and release

Hi Terry,

Quoted 15 lines · Terry Bowman

EFCH cd6h/cd7h port I/O may no longer be available on later AMD processors and it is recommended to use MMIO instead. Update the request and release functions to support MMIO.

MMIO request/release and mmapping require details during cleanup. Add a MMIO configuration structure containing resource and vaddress details for mapping the region, accessing the region, and releasing the region.

---
 drivers/i2c/busses/i2c-piix4.c | 61 ++++++++++++++++++++++++++++++----
 1 file changed, 55 insertions(+), 6 deletions(-)

(...)

drivers/i2c/busses/i2c-piix4.c static int piix4_add_adapter(struct pci_dev *dev, unsigned short smba, · line 931

Quoted 4 lines
drivers/i2c/busses/i2c-piix4.c +1 -0
@@ -880,6 +928,7 @@ static int piix4_add_adapter(struct pci_dev *dev, unsigned short smba,
 		return -ENOMEM;
 	}
 
+	adapdata->mmio_cfg.use_mmio = 0;

Useless initialization, as the adapdata structure has just been zalloc'd.

Quoted 3 lines
drivers/i2c/busses/i2c-piix4.c +0 -0
 	adapdata->smba = smba;
 	adapdata->sb800_main = sb800_main;
 	adapdata->port = port << piix4_port_shift_sb800;
Signature · 2 lines
Jean Delvare
SUSE L3 Support
#12 lore raw

PATCH v4 7/9 i2c: piix4: Add EFCH MMIO support to SMBus base address detect

In reply to Terry Bowman

The EFCH SMBus controller's base address is determined using details in FCH::PM::DECODEEN[smbusasfiobase] and FCH::PM::DECODEEN[smbusasfioen].These register fields were accessed using cd6h/cd7h port I/O. cd6h/cd7h port I/O is no longer available in later AMD processors. Change base address detection to use MMIO instead of port I/O cd6h/cd7h.

---
 drivers/i2c/busses/i2c-piix4.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
drivers/i2c/busses/i2c-piix4.c +9 -4
index 7defa0c5f1f9..a23c0327e1f6 100644
@@ -344,10 +344,15 @@ static int piix4_setup_sb800_smba(struct pci_dev *PIIX4_dev,
 	if (retval)
 		return retval;
 
-	outb_p(smb_en, SB800_PIIX4_SMB_IDX);
-	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
-	outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
-	smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
+	if (mmio_cfg.use_mmio) {
+		smba_en_lo = ioread8(mmio_cfg.addr);
+		smba_en_hi = ioread8(mmio_cfg.addr + 1);
+	} else {
+		outb_p(smb_en, SB800_PIIX4_SMB_IDX);
+		smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
+		outb_p(smb_en + 1, SB800_PIIX4_SMB_IDX);
+		smba_en_hi = inb_p(SB800_PIIX4_SMB_IDX + 1);
+	}
 
 	piix4_sb800_region_release(&PIIX4_dev->dev, &mmio_cfg);
 
Signature · 1 line
2.30.2
#13 lore raw

PATCH v4 8/9 i2c: piix4: Add EFCH MMIO support for SMBus port select

In reply to Terry Bowman

AMD processors include registers capable of selecting between 2 SMBus ports. Port selection is made during each user access by writing to FCH::PM::DECODEEN[smbus0sel]. Change the driver to use MMIO during SMBus port selection because cd6h/cd7h port I/O is not available on later AMD processors.

---
 drivers/i2c/busses/i2c-piix4.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)
drivers/i2c/busses/i2c-piix4.c +13 -3
index a23c0327e1f6..c5325cadaf55 100644
@@ -752,10 +752,19 @@ static void piix4_imc_wakeup(void)
 	release_region(KERNCZ_IMC_IDX, 2);
 }
 
-static int piix4_sb800_port_sel(u8 port)
+static int piix4_sb800_port_sel(u8 port, struct sb800_mmio_cfg *mmio_cfg)
 {
 	u8 smba_en_lo, val;
 
+	if (mmio_cfg->use_mmio) {
+		smba_en_lo = ioread8(mmio_cfg->addr + piix4_port_sel_sb800);
+		val = (smba_en_lo & ~piix4_port_mask_sb800) | port;
+		if (smba_en_lo != val)
+			iowrite8(val, mmio_cfg->addr + piix4_port_sel_sb800);
+
+		return (smba_en_lo & piix4_port_mask_sb800);
+	}
+
 	outb_p(piix4_port_sel_sb800, SB800_PIIX4_SMB_IDX);
 	smba_en_lo = inb_p(SB800_PIIX4_SMB_IDX + 1);
 
@@ -765,6 +774,7 @@ static int piix4_sb800_port_sel(u8 port)
 
 	return (smba_en_lo & piix4_port_mask_sb800);
 }
+
 /*
  * Handles access to multiple SMBus ports on the SB800.
  * The port is selected by bits 2:1 of the smb_en register (0x2c).
@@ -841,12 +851,12 @@ static s32 piix4_access_sb800(struct i2c_adapter *adap, u16 addr,
 		}
 	}
 
-	prev_port = piix4_sb800_port_sel(adapdata->port);
+	prev_port = piix4_sb800_port_sel(adapdata->port, &adapdata->mmio_cfg);
 
 	retval = piix4_access(adap, addr, flags, read_write,
 			      command, size, data);
 
-	piix4_sb800_port_sel(prev_port);
+	piix4_sb800_port_sel(prev_port, &adapdata->mmio_cfg);
 
 	/* Release the semaphore */
 	outb_p(smbslvcnt | 0x20, SMBSLVCNT);
Signature · 1 line
2.30.2
#14 lore raw

Re: [PATCH v4 8/9] i2c: piix4: Add EFCH MMIO support for SMBus port select

Hi Terry,

Quoted 13 lines · Terry Bowman

AMD processors include registers capable of selecting between 2 SMBus ports. Port selection is made during each user access by writing to FCH::PM::DECODEEN[smbus0sel]. Change the driver to use MMIO during SMBus port selection because cd6h/cd7h port I/O is not available on later AMD processors.

---
 drivers/i2c/busses/i2c-piix4.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

(...)

drivers/i2c/busses/i2c-piix4.c static int piix4_sb800_port_sel(u8 port) · line 780

Quoted 7 lines
drivers/i2c/busses/i2c-piix4.c +1 -0
@@ -765,6 +774,7 @@ static int piix4_sb800_port_sel(u8 port)
 
 	return (smba_en_lo & piix4_port_mask_sb800);
 }
+
 /*
  * Handles access to multiple SMBus ports on the SB800.
  * The port is selected by bits 2:1 of the smb_en register (0x2c).

We indeed want a blank line here, but it should be inserted in patch [5/9] (which adds function piix4_sb800_port_sel), not in this patch.

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#15 lore raw

PATCH v4 9/9 i2c: piix4: Enable EFCH MMIO for Family 17h+

In reply to Terry Bowman

Enable EFCH MMIO using check for SMBus PCI revision ID value 0x51 or greater. SMBus PCI revision ID 0x51 is first used by family 17h. This PCI revision ID check will also enable future AMD processors with the same EFCH SMBus controller HW.

---
 drivers/i2c/busses/i2c-piix4.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-piix4.c +11 -2
index c5325cadaf55..6a9495d994bc 100644
@@ -101,6 +101,8 @@
 #define SB800_PIIX4_FCH_PM_ADDR			0xFED80300
 #define SB800_PIIX4_FCH_PM_SIZE			8
 
+#define AMD_PCI_SMBUS_REVISION_MMIO		0x51
+
 /* insmod parameters */
 
 /* If force is set to anything different from 0, we forcibly enable the
@@ -229,6 +231,13 @@ static void piix4_sb800_region_release(struct device *dev,
 	release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 }
 
+static bool piix4_sb800_use_mmio(struct pci_dev *PIIX4_dev)
+{
+	return (PIIX4_dev->vendor == PCI_VENDOR_ID_AMD &&
+		PIIX4_dev->device == PCI_DEVICE_ID_AMD_KERNCZ_SMBUS &&
+		PIIX4_dev->revision >= AMD_PCI_SMBUS_REVISION_MMIO);
+}
+
 static int piix4_setup(struct pci_dev *PIIX4_dev,
 		       const struct pci_device_id *id)
 {
@@ -339,7 +348,7 @@ static int piix4_setup_sb800_smba(struct pci_dev *PIIX4_dev,
 	u8 smba_en_hi;
 	int retval;
 
-	mmio_cfg.use_mmio = 0;
+	mmio_cfg.use_mmio = piix4_sb800_use_mmio(PIIX4_dev);
 	retval = piix4_sb800_region_request(&PIIX4_dev->dev, &mmio_cfg);
 	if (retval)
 		return retval;
@@ -943,7 +952,7 @@ static int piix4_add_adapter(struct pci_dev *dev, unsigned short smba,
 		return -ENOMEM;
 	}
 
-	adapdata->mmio_cfg.use_mmio = 0;
+	adapdata->mmio_cfg.use_mmio = piix4_sb800_use_mmio(dev);
 	adapdata->smba = smba;
 	adapdata->sb800_main = sb800_main;
 	adapdata->port = port << piix4_port_shift_sb800;
Signature · 1 line
2.30.2
#16 lore raw

Re: [PATCH v4 9/9] i2c: piix4: Enable EFCH MMIO for Family 17h+

Hi Terry,

Quoted 15 lines · Terry Bowman

Enable EFCH MMIO using check for SMBus PCI revision ID value 0x51 or greater. SMBus PCI revision ID 0x51 is first used by family 17h. This PCI revision ID check will also enable future AMD processors with the same EFCH SMBus controller HW.

---
 drivers/i2c/busses/i2c-piix4.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-piix4.c +0 -0
index c5325cadaf55..6a9495d994bc 100644

drivers/i2c/busses/i2c-piix4.c line 105

Quoted 5 lines
drivers/i2c/busses/i2c-piix4.c +2 -0
@@ -101,6 +101,8 @@
 #define SB800_PIIX4_FCH_PM_ADDR			0xFED80300
 #define SB800_PIIX4_FCH_PM_SIZE			8
 
+#define AMD_PCI_SMBUS_REVISION_MMIO		0x51
+

I don't think that was worth a define. You only use the value once, in a context where the symbolic name doesn't add much value IMHO.

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#17 lore raw

Re: [PATCH v4 9/9] i2c: piix4: Enable EFCH MMIO for Family 17h+

Quoted 1 line · Jean Delvare

On Sun, 30 Jan 2022 12:41:30 -0600, Terry Bowman wrote:

...

Quoted 4 lines
Quoted 1 line

+#define AMD_PCI_SMBUS_REVISION_MMIO 0x51

I don't think that was worth a define. You only use the value once, in a context where the symbolic name doesn't add much value IMHO.

I don't remember the code context here, but it would be nice in such a case to convert this definition to a comment (if it's not crystal clear what this magic number is about).

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#18 lore raw

Re: [PATCH v4 9/9] i2c: piix4: Enable EFCH MMIO for Family 17h+

Hi Andy,

Quoted 5 lines · Andy Shevchenko
Quoted 1 line · Jean Delvare

On Sun, 30 Jan 2022 12:41:30 -0600, Terry Bowman wrote:

...

Quoted 10 lines
Quoted 4 lines
Quoted 1 line

+#define AMD_PCI_SMBUS_REVISION_MMIO 0x51

I don't think that was worth a define. You only use the value once, in a context where the symbolic name doesn't add much value IMHO.

I don't remember the code context here, but it would be nice in such a case to convert this definition to a comment (if it's not crystal clear what this magic number is about).

drivers/i2c/busses/i2c-piix4.c +7 -0
index 239df17ce02b..218ed8efb83e 100644
@@ -229,6 +229,13 @@ static void piix4_sb800_region_release(struct device *dev,
        release_region(SB800_PIIX4_SMB_IDX, SB800_PIIX4_SMB_MAP_SIZE);
 }
 
+static bool piix4_sb800_use_mmio(struct pci_dev *PIIX4_dev)
+{
+       return (PIIX4_dev->vendor == PCI_VENDOR_ID_AMD &&
+               PIIX4_dev->device == PCI_DEVICE_ID_AMD_KERNCZ_SMBUS &&
+               PIIX4_dev->revision >= AMD_PCI_SMBUS_REVISION_MMIO);
+}
+
 static int piix4_setup(struct pci_dev *PIIX4_dev,
                       const struct pci_device_id *id)
 {

----

I added the context above. I'll add a comment in v5 to describe what and why 0x51 is used.

Regards,
Terry

Reviewed-by on the v4 series
#19 lore raw
In reply to Terry Bowman
Quoted 35 lines · Terry Bowman

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

This series includes patches with MMIO accesses to register FCH::PM::DECODEEN. The same register is also accessed by the sp5100_tco driver.[1] Synchronization to the MMIO register is required in both drivers.

The first patch creates a macro to request MMIO region using the 'muxed' retry logic. This is used in patch 6 to synchronize accesses to EFCH MMIO.

The second patch replaces a hardcoded region size with a #define. This is to improve maintainability and was requested from v2 review.

The third patch moves duplicated region request/release code into functions. This locates related code into functions and reduces code line count. This will also make adding MMIO support in patch 6 easier.

The fourth patch moves SMBus controller address detection into a function. This is in preparation for adding MMIO region support.

The fifth patch moves EFCH port selection into a function. This is in preparation for adding MMIO region support.

The sixth patch adds MMIO support for region requesting/releasing and mapping. This is necessary for using MMIO to detect SMBus controller address, enable SMBbus controller region, and control the port select.

The seventh patch updates the SMBus controller address detection to support using MMIO. This is necessary because the driver accesses register FCH::PM::DECODEEN during initialization and only available using MMIO on later AMD processors.

Quoted 8 lines

The eighth patch updates the SMBus port selection to support MMIO. This is required because port selection control resides in the FCH::PM::DECODEEN[smbus0sel] and is only accessible using MMIO on later AMD processors.

The ninth patch enables the EFCH MMIO functionality added earlier in this series. The SMBus controller's PCI revision ID is used to check if EFCH MMIO is supported by HW and should be enabled in the driver.

I'm not against the series, but I am still concerned that we are using _p IO without clear understanding why. With cleaning them up, this can be simpler and cleaner.

For the i2c patches:

if it helps somebody.

Quoted 1 line

Based on v5.16.

v5.17-rc2 is out :-)

Quoted 95 lines

Testing:
Tested on family 19h using:
i2cdetect -y 0
i2cdetect -y 2

- Results using v5.16 and this pachset applied. Below

    shows the devices detected on the busses:

    # i2cdetect -y 0
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- --
    10: 10 11 -- -- -- -- -- -- 18 -- -- -- -- -- -- --
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    30: 30 -- -- -- -- 35 36 -- -- -- -- -- -- -- -- --
    40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- --
    50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    70: -- -- -- 73 -- -- -- --
    # i2cdetect -y 2
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- --
    10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- --
    50: -- 51 -- -- 54 -- -- -- -- -- -- -- -- -- -- --
    60: 60 -- -- 63 -- -- 66 -- -- -- -- 6b -- -- 6e --
    70: 70 71 72 73 74 75 -- 77

Also tested using sp5100_tco submitted series listed below.[1] I applied the sp5100_tco v4 series and ran:
cat >> /dev/watchdog

[1] sp5100_tco v4 patchset is in process but not sent yet. Below is v3 upstream review:

Changes in v4:
- Changed request_muxed_mem_region() macro to request_mem_region_muxed() in patch 1. (Andy Shevchenko)
- Removed unnecessary newline where possible in calls to request_muxed_region() in patch 2. (Terry Bowman)
- Changed piix4_sb800_region_setup() to piix4_sb800_region_request(). Patch 3. (Jean Delvare)
- Reordered piix4_setup_sb800() local variables from longest name length to shortest name length. Patch 4. (Terry Bowman)
- Changed piix4_sb800_region_request() and piix4_sb800_region_release() by adding early return() to remove 'else' improving readability. Patch 6. (Terry Bowman)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)
- Refactored piix4_sb800_port_sel() to simplify the 'if' statement using temp variable. Patch 8. (Terry Bowman)
- Added mmio_cfg.use_mmio assignment in piix4_add_adapter(). This is needed for calls to piix4_sb800_port_sel() after initialization during normal operation. Patch 9. (Terry Bowman)

Changes in v3:
- Added request_muxed_mem_region() patch (Wolfram, Guenter)
- Reduced To/Cc list length. (Andy)

Changes in v2:
- Split single patch. (Jean Delvare)
- Replace constant 2 with SB800_PIIX4_SMB_MAP_SIZE where appropriate. (Jean Delvare)
- Shorten SB800_PIIX4_FCH_PM_DECODEEN_MMIO_EN name length to SB800_PIIX4_FCH_PM_DECODEEN_MMIO. (Jean Delvare)
- Change AMD_PCI_SMBUS_REVISION_MMIO from 0x59 to 0x51. (Terry Bowman)
- Change piix4_sb800_region_setup() to piix4_sb800_region_request(). (Jean Delvare)
- Change 'SMB' text in logging to 'SMBus' (Jean Delvare)
- Remove unnecessary NULL assignment in piix4_sb800_region_release(). (Jean Delvare)
- Move 'u8' variable definitions to single line. (Jean Delvare)
- Hardcode piix4_setup_sb800_smba() return value to 0 since it is always
0. (Jean Delvare)

Terry Bowman (9):
kernel/resource: Introduce request_mem_region_muxed()
i2c: piix4: Replace hardcoded memory map size with a #define
i2c: piix4: Move port I/O region request/release code into functions
i2c: piix4: Move SMBus controller base address detect into function
i2c: piix4: Move SMBus port selection into function
i2c: piix4: Add EFCH MMIO support to region request and release
i2c: piix4: Add EFCH MMIO support to SMBus base address detect
i2c: piix4: Add EFCH MMIO support for SMBus port select
i2c: piix4: Enable EFCH MMIO for Family 17h+

 drivers/i2c/busses/i2c-piix4.c | 207 ++++++++++++++++++++++++++-------
 include/linux/ioport.h         |   2 +
 2 files changed, 164 insertions(+), 45 deletions(-)

--
2.30.2

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#20 lore raw

Hi Andy,

Quoted 45 lines · Andy Shevchenko
Quoted 35 lines · Terry Bowman

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

This series includes patches with MMIO accesses to register FCH::PM::DECODEEN. The same register is also accessed by the sp5100_tco driver.[1] Synchronization to the MMIO register is required in both drivers.

The first patch creates a macro to request MMIO region using the 'muxed' retry logic. This is used in patch 6 to synchronize accesses to EFCH MMIO.

The second patch replaces a hardcoded region size with a #define. This is to improve maintainability and was requested from v2 review.

The third patch moves duplicated region request/release code into functions. This locates related code into functions and reduces code line count. This will also make adding MMIO support in patch 6 easier.

The fourth patch moves SMBus controller address detection into a function. This is in preparation for adding MMIO region support.

The fifth patch moves EFCH port selection into a function. This is in preparation for adding MMIO region support.

The sixth patch adds MMIO support for region requesting/releasing and mapping. This is necessary for using MMIO to detect SMBus controller address, enable SMBbus controller region, and control the port select.

The seventh patch updates the SMBus controller address detection to support using MMIO. This is necessary because the driver accesses register FCH::PM::DECODEEN during initialization and only available using MMIO on later AMD processors.

Quoted 8 lines

The eighth patch updates the SMBus port selection to support MMIO. This is required because port selection control resides in the FCH::PM::DECODEEN[smbus0sel] and is only accessible using MMIO on later AMD processors.

The ninth patch enables the EFCH MMIO functionality added earlier in this series. The SMBus controller's PCI revision ID is used to check if EFCH MMIO is supported by HW and should be enabled in the driver.

Quoted 8 lines

I'm not against the series, but I am still concerned that we are using _p IO without clear understanding why. With cleaning them up, this can be simpler and cleaner.

For the i2c patches:

if it helps somebody.

Thanks for the review Andy! I plan to add the ioport_map changes you recommend in a future patchset. I will include you as "suggested-by".

Regards,
Terry

Quoted 102 lines
Quoted 1 line

Based on v5.16.

v5.17-rc2 is out :-)

Quoted 95 lines

Testing:
Tested on family 19h using:
i2cdetect -y 0
i2cdetect -y 2

- Results using v5.16 and this pachset applied. Below

    shows the devices detected on the busses:

    # i2cdetect -y 0
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- --
    10: 10 11 -- -- -- -- -- -- 18 -- -- -- -- -- -- --
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    30: 30 -- -- -- -- 35 36 -- -- -- -- -- -- -- -- --
    40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- --
    50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    70: -- -- -- 73 -- -- -- --
    # i2cdetect -y 2
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- --
    10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- --
    50: -- 51 -- -- 54 -- -- -- -- -- -- -- -- -- -- --
    60: 60 -- -- 63 -- -- 66 -- -- -- -- 6b -- -- 6e --
    70: 70 71 72 73 74 75 -- 77

Also tested using sp5100_tco submitted series listed below.[1] I applied the sp5100_tco v4 series and ran:
cat >> /dev/watchdog

[1] sp5100_tco v4 patchset is in process but not sent yet. Below is v3 upstream review:

Changes in v4:
- Changed request_muxed_mem_region() macro to request_mem_region_muxed() in patch 1. (Andy Shevchenko)
- Removed unnecessary newline where possible in calls to request_muxed_region() in patch 2. (Terry Bowman)
- Changed piix4_sb800_region_setup() to piix4_sb800_region_request(). Patch 3. (Jean Delvare)
- Reordered piix4_setup_sb800() local variables from longest name length to shortest name length. Patch 4. (Terry Bowman)
- Changed piix4_sb800_region_request() and piix4_sb800_region_release() by adding early return() to remove 'else' improving readability. Patch 6. (Terry Bowman)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)
- Refactored piix4_sb800_port_sel() to simplify the 'if' statement using temp variable. Patch 8. (Terry Bowman)
- Added mmio_cfg.use_mmio assignment in piix4_add_adapter(). This is needed for calls to piix4_sb800_port_sel() after initialization during normal operation. Patch 9. (Terry Bowman)

Changes in v3:
- Added request_muxed_mem_region() patch (Wolfram, Guenter)
- Reduced To/Cc list length. (Andy)

Changes in v2:
- Split single patch. (Jean Delvare)
- Replace constant 2 with SB800_PIIX4_SMB_MAP_SIZE where appropriate. (Jean Delvare)
- Shorten SB800_PIIX4_FCH_PM_DECODEEN_MMIO_EN name length to SB800_PIIX4_FCH_PM_DECODEEN_MMIO. (Jean Delvare)
- Change AMD_PCI_SMBUS_REVISION_MMIO from 0x59 to 0x51. (Terry Bowman)
- Change piix4_sb800_region_setup() to piix4_sb800_region_request(). (Jean Delvare)
- Change 'SMB' text in logging to 'SMBus' (Jean Delvare)
- Remove unnecessary NULL assignment in piix4_sb800_region_release(). (Jean Delvare)
- Move 'u8' variable definitions to single line. (Jean Delvare)
- Hardcode piix4_setup_sb800_smba() return value to 0 since it is always
0. (Jean Delvare)

Terry Bowman (9):
kernel/resource: Introduce request_mem_region_muxed()
i2c: piix4: Replace hardcoded memory map size with a #define
i2c: piix4: Move port I/O region request/release code into functions
i2c: piix4: Move SMBus controller base address detect into function
i2c: piix4: Move SMBus port selection into function
i2c: piix4: Add EFCH MMIO support to region request and release
i2c: piix4: Add EFCH MMIO support to SMBus base address detect
i2c: piix4: Add EFCH MMIO support for SMBus port select
i2c: piix4: Enable EFCH MMIO for Family 17h+

 drivers/i2c/busses/i2c-piix4.c | 207 ++++++++++++++++++++++++++-------
 include/linux/ioport.h         |   2 +
 2 files changed, 164 insertions(+), 45 deletions(-)

--
2.30.2

#21 lore raw
In reply to Terry Bowman
Quoted 3 lines · Terry Bowman

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

Review from Andy is already great, I'll give Jean a few more days for comments.

Reviewed-by on the v4 series
#22 lore raw
In reply to Terry Bowman

Hi Terry,

Quoted 34 lines · Terry Bowman

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

This series includes patches with MMIO accesses to register FCH::PM::DECODEEN. The same register is also accessed by the sp5100_tco driver.[1] Synchronization to the MMIO register is required in both drivers.

The first patch creates a macro to request MMIO region using the 'muxed' retry logic. This is used in patch 6 to synchronize accesses to EFCH MMIO.

The second patch replaces a hardcoded region size with a #define. This is to improve maintainability and was requested from v2 review.

The third patch moves duplicated region request/release code into functions. This locates related code into functions and reduces code line count. This will also make adding MMIO support in patch 6 easier.

The fourth patch moves SMBus controller address detection into a function. This is in preparation for adding MMIO region support.

The fifth patch moves EFCH port selection into a function. This is in preparation for adding MMIO region support.

The sixth patch adds MMIO support for region requesting/releasing and mapping. This is necessary for using MMIO to detect SMBus controller address, enable SMBbus controller region, and control the port select.

The seventh patch updates the SMBus controller address detection to support using MMIO. This is necessary because the driver accesses register FCH::PM::DECODEEN during initialization and only available using MMIO on later AMD processors.

Quoted 8 lines

The eighth patch updates the SMBus port selection to support MMIO. This is required because port selection control resides in the FCH::PM::DECODEEN[smbus0sel] and is only accessible using MMIO on later AMD processors.

The ninth patch enables the EFCH MMIO functionality added earlier in this series. The SMBus controller's PCI revision ID is used to check if EFCH MMIO is supported by HW and should be enabled in the driver.

Thank you for splitting the changes into small chunks for easier review. Maybe it was even a bit too much in the end, as most patches don't serve a purpose on their own. But well, that's still much better than a monolithic patch.

Quoted 22 lines

Based on v5.16.

Testing:
Tested on family 19h using:
i2cdetect -y 0
i2cdetect -y 2

- Results using v5.16 and this pachset applied. Below

    shows the devices detected on the busses:
    
    # i2cdetect -y 0 
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- -- 
    10: 10 11 -- -- -- -- -- -- 18 -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: 30 -- -- -- -- 35 36 -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- -- 
    50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    70: -- -- -- 73 -- -- -- --                         
    # i2cdetect -y 2
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
Quoted 8 lines
    00:                         -- -- -- -- -- -- -- -- 
    10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- -- 
    50: -- 51 -- -- 54 -- -- -- -- -- -- -- -- -- -- -- 
    60: 60 -- -- 63 -- -- 66 -- -- -- -- 6b -- -- 6e -- 
    70: 70 71 72 73 74 75 -- 77

Unfortunately I'm not really able to test this series. While I do have an AMD-based laptop which uses the i2c-piix4 driver, the SMBus has never been usable on it. The driver creates 3 i2c buses (port 0 at 0b00, port 2 at 0b00 and port 1 at 0b20). The first 2 do not seem to have any device on them (i2cdetect returns empty). The last one must have some devices on it, I see address 0x1c answers the ping, unfortunately as soon as probing reaches address 0x2c, all later pings return success, regardless of the address. It seems that some I2C device (probably the one at 0x2c, but I don't know what it is) is holding SDA low forever, therefore preventing any use of the whole SMBus port until the next reboot.

Quoted 5 lines

(...)
Changes in v4:
(...)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)

I'm curious, how can you be sure of that actually?

Quoted 12 lines

(...)
Terry Bowman (9):
kernel/resource: Introduce request_mem_region_muxed()
i2c: piix4: Replace hardcoded memory map size with a #define
i2c: piix4: Move port I/O region request/release code into functions
i2c: piix4: Move SMBus controller base address detect into function
i2c: piix4: Move SMBus port selection into function
i2c: piix4: Add EFCH MMIO support to region request and release
i2c: piix4: Add EFCH MMIO support to SMBus base address detect
i2c: piix4: Add EFCH MMIO support for SMBus port select
i2c: piix4: Enable EFCH MMIO for Family 17h+

Quoted 3 lines
 drivers/i2c/busses/i2c-piix4.c | 207 ++++++++++++++++++++++++++-------
 include/linux/ioport.h         |   2 +
 2 files changed, 164 insertions(+), 45 deletions(-)

I'm done with my review, looks good overall, I made a few comments here and there but no major issue. I'll leave it up to you (and Wolfram) to either send a new series with (some of) my suggestions addressed or just go with v4. In both cases you can add:

to all patches.

Thank you very much for your work, and sorry for my late review.

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#23 lore raw

Hi Jean,

Quoted 82 lines · Jean Delvare

Hi Terry,

Quoted 34 lines · Terry Bowman

This series changes the piix4_smbus driver's cd6h/cd7h port I/O accesses to use MMIO instead. This is necessary because cd6h/cd7h port I/O may be disabled on later AMD processors.

This series includes patches with MMIO accesses to register FCH::PM::DECODEEN. The same register is also accessed by the sp5100_tco driver.[1] Synchronization to the MMIO register is required in both drivers.

The first patch creates a macro to request MMIO region using the 'muxed' retry logic. This is used in patch 6 to synchronize accesses to EFCH MMIO.

The second patch replaces a hardcoded region size with a #define. This is to improve maintainability and was requested from v2 review.

The third patch moves duplicated region request/release code into functions. This locates related code into functions and reduces code line count. This will also make adding MMIO support in patch 6 easier.

The fourth patch moves SMBus controller address detection into a function. This is in preparation for adding MMIO region support.

The fifth patch moves EFCH port selection into a function. This is in preparation for adding MMIO region support.

The sixth patch adds MMIO support for region requesting/releasing and mapping. This is necessary for using MMIO to detect SMBus controller address, enable SMBbus controller region, and control the port select.

The seventh patch updates the SMBus controller address detection to support using MMIO. This is necessary because the driver accesses register FCH::PM::DECODEEN during initialization and only available using MMIO on later AMD processors.

Quoted 8 lines

The eighth patch updates the SMBus port selection to support MMIO. This is required because port selection control resides in the FCH::PM::DECODEEN[smbus0sel] and is only accessible using MMIO on later AMD processors.

The ninth patch enables the EFCH MMIO functionality added earlier in this series. The SMBus controller's PCI revision ID is used to check if EFCH MMIO is supported by HW and should be enabled in the driver.

Thank you for splitting the changes into small chunks for easier review. Maybe it was even a bit too much in the end, as most patches don't serve a purpose on their own. But well, that's still much better than a monolithic patch.

Quoted 22 lines

Based on v5.16.

Testing:
Tested on family 19h using:
i2cdetect -y 0
i2cdetect -y 2

- Results using v5.16 and this pachset applied. Below

    shows the devices detected on the busses:
    
    # i2cdetect -y 0 
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
    00:                         -- -- -- -- -- -- -- -- 
    10: 10 11 -- -- -- -- -- -- 18 -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: 30 -- -- -- -- 35 36 -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- -- 
    50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    70: -- -- -- 73 -- -- -- --                         
    # i2cdetect -y 2
         0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
Quoted 8 lines
    00:                         -- -- -- -- -- -- -- -- 
    10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
    40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- -- 
    50: -- 51 -- -- 54 -- -- -- -- -- -- -- -- -- -- -- 
    60: 60 -- -- 63 -- -- 66 -- -- -- -- 6b -- -- 6e -- 
    70: 70 71 72 73 74 75 -- 77
Quoted 12 lines

Unfortunately I'm not really able to test this series. While I do have an AMD-based laptop which uses the i2c-piix4 driver, the SMBus has never been usable on it. The driver creates 3 i2c buses (port 0 at 0b00, port 2 at 0b00 and port 1 at 0b20). The first 2 do not seem to have any device on them (i2cdetect returns empty). The last one must have some devices on it, I see address 0x1c answers the ping, unfortunately as soon as probing reaches address 0x2c, all later pings return success, regardless of the address. It seems that some I2C device (probably the one at 0x2c, but I don't know what it is) is holding SDA low forever, therefore preventing any use of the whole SMBus port until the next reboot.

My understanding is the OEM may have different i2c devices because it is mainboard specific.

Quoted 8 lines
Quoted 5 lines

(...)
Changes in v4:
(...)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)

I'm curious, how can you be sure of that actually?

The removed code was using a MMIO region to write 1 to 'mmioen'. This was using the MMIO region to enable same MMIO region.

Programmer's manual shows FCH::PM::ISACONTROL[mmioen] has a '1' reset value. Per programmer's manual, FCH::PM::ISACONTROL[mmioen] enables MMIO mapping at FED8_0000h..FED8_1FFFh. The FCH::PM::ISACONTROL register is MMIO mapped at FED8_0304h. 'mmioen' was intended to be set using port I/O to enable MMIO but, port I/O access to these registers is now disabled.

Quoted 12 lines
Quoted 12 lines

(...)
Terry Bowman (9):
kernel/resource: Introduce request_mem_region_muxed()
i2c: piix4: Replace hardcoded memory map size with a #define
i2c: piix4: Move port I/O region request/release code into functions
i2c: piix4: Move SMBus controller base address detect into function
i2c: piix4: Move SMBus port selection into function
i2c: piix4: Add EFCH MMIO support to region request and release
i2c: piix4: Add EFCH MMIO support to SMBus base address detect
i2c: piix4: Add EFCH MMIO support for SMBus port select
i2c: piix4: Enable EFCH MMIO for Family 17h+

Quoted 9 lines
Quoted 3 lines
 drivers/i2c/busses/i2c-piix4.c | 207 ++++++++++++++++++++++++++-------
 include/linux/ioport.h         |   2 +
 2 files changed, 164 insertions(+), 45 deletions(-)

I'm done with my review, looks good overall, I made a few comments here and there but no major issue. I'll leave it up to you (and Wolfram) to either send a new series with (some of) my suggestions addressed or just go with v4. In both cases you can add:

I'll add your requested fixes into v5 and will send for review this afternoon.

Quoted 6 lines

to all patches.

Thank you very much for your work, and sorry for my late review.

Thanks for the review.

Regards,
Terry

#24 lore raw
Quoted 13 lines · Terry Bowman
Quoted 11 lines · Jean Delvare

Unfortunately I'm not really able to test this series. While I do have an AMD-based laptop which uses the i2c-piix4 driver, the SMBus has never been usable on it. The driver creates 3 i2c buses (port 0 at 0b00, port 2 at 0b00 and port 1 at 0b20). The first 2 do not seem to have any device on them (i2cdetect returns empty). The last one must have some devices on it, I see address 0x1c answers the ping, unfortunately as soon as probing reaches address 0x2c, all later pings return success, regardless of the address. It seems that some I2C device (probably the one at 0x2c, but I don't know what it is) is holding SDA low forever, therefore preventing any use of the whole SMBus port until the next reboot.

Quoted 2 lines

My understanding is the OEM may have different i2c devices because it is mainboard specific.

Yes, the devices on the SMBus could be anything Lenovo decided to use. Tomorrow I'll try to scan the SMBus but skipping the problematic address, to see if it works around the problem.

Quoted 10 lines
Quoted 7 lines
Quoted 5 lines

(...)
Changes in v4:
(...)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)

I'm curious, how can you be sure of that actually?

The removed code was using a MMIO region to write 1 to 'mmioen'. This was using the MMIO region to enable same MMIO region.

Ah ah, I get it now. Nice chicken or the egg situation :-)

Quoted 5 lines

Programmer's manual shows FCH::PM::ISACONTROL[mmioen] has a '1' reset value. Per programmer's manual, FCH::PM::ISACONTROL[mmioen] enables MMIO mapping at FED8_0000h..FED8_1FFFh. The FCH::PM::ISACONTROL register is MMIO mapped at FED8_0304h. 'mmioen' was intended to be set using port I/O to enable MMIO but, port I/O access to these registers is now disabled.

Is my understanding correct that there is some overlapping of the access methods and there are certain chipsets where both legacy I/O and MMIO access is possible?

If so, while there's indeed nothing to be done for the most recent systems where only MMIO access is possible, you may still need to enable MMIO access through legacy I/O if you try to use MMIO on chipsets where both are possible. I'm not sure what exactly where you set the limit. In the last patch you say that 0x51 is the first revision of the family 17h CPUs, but is family 17h the first where MMIO is available, or the first where legacy I/O isn't?

Signature · 2 lines
Jean Delvare
SUSE L3 Support
#25 lore raw

Hi Jean,

Quoted 40 lines · Jean Delvare
Quoted 13 lines · Terry Bowman
Quoted 11 lines · Jean Delvare

Unfortunately I'm not really able to test this series. While I do have an AMD-based laptop which uses the i2c-piix4 driver, the SMBus has never been usable on it. The driver creates 3 i2c buses (port 0 at 0b00, port 2 at 0b00 and port 1 at 0b20). The first 2 do not seem to have any device on them (i2cdetect returns empty). The last one must have some devices on it, I see address 0x1c answers the ping, unfortunately as soon as probing reaches address 0x2c, all later pings return success, regardless of the address. It seems that some I2C device (probably the one at 0x2c, but I don't know what it is) is holding SDA low forever, therefore preventing any use of the whole SMBus port until the next reboot.

Quoted 2 lines

My understanding is the OEM may have different i2c devices because it is mainboard specific.

Yes, the devices on the SMBus could be anything Lenovo decided to use. Tomorrow I'll try to scan the SMBus but skipping the problematic address, to see if it works around the problem.

Quoted 10 lines
Quoted 7 lines
Quoted 5 lines

(...)
Changes in v4:
(...)
- Removed iowrite32(ioread32(...), ...). Unnecessary because MMIO is already enabled. (Terry Bowman)

I'm curious, how can you be sure of that actually?

The removed code was using a MMIO region to write 1 to 'mmioen'. This was using the MMIO region to enable same MMIO region.

Ah ah, I get it now. Nice chicken or the egg situation :-)

Quoted 5 lines

Programmer's manual shows FCH::PM::ISACONTROL[mmioen] has a '1' reset value. Per programmer's manual, FCH::PM::ISACONTROL[mmioen] enables MMIO mapping at FED8_0000h..FED8_1FFFh. The FCH::PM::ISACONTROL register is MMIO mapped at FED8_0304h. 'mmioen' was intended to be set using port I/O to enable MMIO but, port I/O access to these registers is now disabled.

Quoted 4 lines

Is my understanding correct that there is some overlapping of the access methods and there are certain chipsets where both legacy I/O and MMIO access is possible?

Yes.

Quoted 8 lines

If so, while there's indeed nothing to be done for the most recent systems where only MMIO access is possible, you may still need to enable MMIO access through legacy I/O if you try to use MMIO on chipsets where both are possible. I'm not sure what exactly where you set the limit. In the last patch you say that 0x51 is the first revision of the family 17h CPUs, but is family 17h the first where MMIO is available, or the first where legacy I/O isn't?

Family 17h, SMBus PCI ID >= 0x51 is the first where cd6h/cd7h port I/O is disabled. If SMBus PCI ID < 0x51 then cd6h/cd7h port I/O is used.

Regards,
Terry

#26 lore raw
Quoted 11 lines · Terry Bowman
Quoted 7 lines · Jean Delvare

If so, while there's indeed nothing to be done for the most recent systems where only MMIO access is possible, you may still need to enable MMIO access through legacy I/O if you try to use MMIO on chipsets where both are possible. I'm not sure what exactly where you set the limit. In the last patch you say that 0x51 is the first revision of the family 17h CPUs, but is family 17h the first where MMIO is available, or the first where legacy I/O isn't?

Family 17h, SMBus PCI ID >= 0x51 is the first where cd6h/cd7h port I/O is disabled. If SMBus PCI ID < 0x51 then cd6h/cd7h port I/O is used.

OK, we are safe then :-)

Signature · 2 lines
Jean Delvare
SUSE L3 Support