net: sfp: add SMBus I2C block support
Subsystems (4)
PATCH v3 1/1 net: sfp: add SMBus I2C block support
Commit 7662abf4db94 ("net: phy: sfp: Add support for SMBus module access") added support for SMBus-only controllers for module access. However, this is restricted to single-byte accesses and has the implication that hwmon is disabled (due to missing atomicity of 16-bit accesses) and warnings are printed.
There are probably a lot of SMBus-only I2C controllers out in the wild which support block reads. Right now, they don't work with SFP modules. This applies - amongst others - to I2C/SMBus-only controllers in Realtek longan and mango SoCs.
Downstream in OpenWrt, a patch similar to the abovementioned patch is used for current LTS kernel 6.12. However, this uses byte-access for all kinds of access and thus disregards the atomicity for wider access.
Introduce read/write SMBus I2C block operations to support SMBus-only controllers with appropriate support for block read/write. Those operations are used for all accesses if supported, otherwise the single-byte operations will be used. With block reads, atomicity for 16-bit reads as required by hwmon is preserved and thus, hwmon can be used.
The implementation requires the I2C_FUNC_SMBUS_I2C_BLOCK to be supported as it relies on reading a pre-defined amount of bytes. This isn't intended by the official SMBus Block Read but supported by several I2C controllers/drivers.
Support for word access is not implemented due to issues regarding endianness.
---
v3: fix previous attempt of v2 to fix return value
v2: return number of written bytes in sfp_smbus_block_write
v2: https://lore.kernel.org/netdev/[email protected]/
v1: https://lore.kernel.org/netdev/[email protected]/
--- drivers/net/phy/sfp.c | 77 +++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 75 insertions(+), 2 deletions(-)
drivers/net/phy/sfp.c +75 -2
@@ -744,6 +744,35 @@ static int sfp_smbus_byte_read(struct sfp *sfp, bool a2, u8 dev_addr,return data - (u8 *)buf; } +static int sfp_smbus_block_read (struct sfp *sfp,bool a2,u8 dev_addr, +void *buf,size_t len) +{ +size_t block_size = sfp->i2c_block_size; +union i2c_smbus_data smbus_data; +u8 bus_addr = a2 ?0x51 :0x50 ; +u8 *data = buf; +u8 this_len; +int ret; + +while (len) { + this_len =min (len, block_size); + + smbus_data.block[0 ] = this_len; + ret =i2c_smbus_xfer (sfp->i2c, bus_addr,0 , +I2C_SMBUS_READ , dev_addr, +I2C_SMBUS_I2C_BLOCK_DATA , &smbus_data); +if (ret <0 ) +return ret; + +memcpy (data, &smbus_data.block[1 ], this_len); + len -= this_len; + data += this_len; + dev_addr += this_len; + } + +return data - (u8 *)buf; +} +static int sfp_smbus_byte_write (struct sfp *sfp,bool a2,u8 dev_addr,void *buf,size_t len) {
@@ -768,23 +797,67 @@ static int sfp_smbus_byte_write(struct sfp *sfp, bool a2, u8 dev_addr,return data - (u8 *)buf; } +static int sfp_smbus_block_write (struct sfp *sfp,bool a2,u8 dev_addr, +void *buf,size_t len) +{ +size_t block_size = sfp->i2c_block_size; +union i2c_smbus_data smbus_data; +u8 bus_addr = a2 ?0x51 :0x50 ; +u8 *data = buf; +u8 this_len; +int ret; + +while (len) { + this_len =min (len, block_size); + + smbus_data.block[0 ] = this_len; +memcpy (&smbus_data.block[1 ], data, this_len); + ret =i2c_smbus_xfer (sfp->i2c, bus_addr,0 , +I2C_SMBUS_WRITE , dev_addr, +I2C_SMBUS_I2C_BLOCK_DATA , &smbus_data); +if (ret) +return ret; + + len -= this_len; + data += this_len; + dev_addr += this_len; + } + +return data - (u8 *)buf; +} +static int sfp_i2c_configure (struct sfp *sfp,struct i2c_adapter *i2c) { +size_t max_block_size; + sfp->i2c = i2c;if (i2c_check_functionality (i2c,I2C_FUNC_I2C )) { sfp->read = sfp_i2c_read; sfp->write = sfp_i2c_write; - sfp->i2c_max_block_size =SFP_EEPROM_BLOCK_SIZE ; + max_block_size =SFP_EEPROM_BLOCK_SIZE ; + }else if (i2c_check_functionality (i2c,I2C_FUNC_SMBUS_I2C_BLOCK )) { + sfp->read = sfp_smbus_block_read; + sfp->write = sfp_smbus_block_write; + + max_block_size =SFP_EEPROM_BLOCK_SIZE ; +if (i2c->quirks && i2c->quirks->max_read_len) + max_block_size =min (max_block_size, + i2c->quirks->max_read_len); +if (i2c->quirks && i2c->quirks->max_write_len) + max_block_size =min (max_block_size, + i2c->quirks->max_write_len); + }else if (i2c_check_functionality (i2c,I2C_FUNC_SMBUS_BYTE_DATA )) { sfp->read = sfp_smbus_byte_read; sfp->write = sfp_smbus_byte_write; - sfp->i2c_max_block_size =1 ; + max_block_size =1 ; }else { sfp->i2c =NULL ;return -EINVAL ; } + sfp->i2c_max_block_size = max_block_size;return 0 ; }
base-commit: c303e8b86d9dbd6868f5216272973292f7f3b7f1
prerequisite-patch-id: ae039dad1e17867fce9182b6b36ac3b1926b254a
Signature · 1 line
2.48.1
Russell King (Oracle)
maintainerQuoted 2 lines · Jonas Jelonek
base-commit:
c303e8b86d9dbd6868f5216272973292f7f3b7f1
prerequisite-patch-id:ae039dad1e17867fce9182b6b36ac3b1926b254a
This seems to be almost useless information. While base-commit exists in the net-next tree, commit ae039dad1e17867fce9182b6b36ac3b1926b254a doesn't exist in either net-next nor net trees.
My guess is you applied Maxime's patch locally, and that is the commit ID of that patch.
Given that Maxime's patch is targetting the net tree (because it's a fix), and your patch is new development, so is for net-next, you either need to:
- wait until Maxime's patch has been merged, and then the net tree has been merged into net-next.
or:
- resend without Maxime's patch.
In either case, please read https://docs.kernel.org/process/maintainer-netdev.html
In particular, the salient points are:
- no resends in under 24 hours (see point 1.6.7)
- specify the tree that your patch is targetting in the subject line (see point 1.6.1) E.g. [PATCH net-next v...] net: blah blah
Thanks.
Signature · 2 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/ FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!
Hi Russell,
Quoted 9 lines · Russell King (Oracle)
Quoted 2 lines · Jonas Jelonek
base-commit:
c303e8b86d9dbd6868f5216272973292f7f3b7f1
prerequisite-patch-id:ae039dad1e17867fce9182b6b36ac3b1926b254aThis seems to be almost useless information. While base-commit exists in the net-next tree, commit
ae039dad1e17867fce9182b6b36ac3b1926b254adoesn't exist in either net-next nor net trees.My guess is you applied Maxime's patch locally, and that is the commit ID of that patch.
This was supposed to be the stable patch-id obtained with 'git patch-id --stable'.
Quoted 8 lines
Given that Maxime's patch is targetting the net tree (because it's a fix), and your patch is new development, so is for net-next, you either need to:
- wait until Maxime's patch has been merged, and then the net tree has been merged into net-next.
or:
- resend without Maxime's patch.
Quoted 8 lines
In either case, please read https://docs.kernel.org/process/maintainer-netdev.html
In particular, the salient points are:
- no resends in under 24 hours (see point 1.6.7)
- specify the tree that your patch is targetting in the subject line (see point 1.6.1) E.g. [PATCH net-next v...] net: blah blahThanks.
Ack. I'll fix the points you mentioned.
Thanks for your patience.
Best,
Jonas
Russell King (Oracle)
maintainerQuoted 10 lines · Jonas Jelonek
Hi Russell,
Quoted 7 lines · Russell King (Oracle)
Quoted 2 lines · Jonas Jelonek
base-commit:
c303e8b86d9dbd6868f5216272973292f7f3b7f1
prerequisite-patch-id:ae039dad1e17867fce9182b6b36ac3b1926b254aThis seems to be almost useless information. While base-commit exists in the net-next tree, commit
ae039dad1e17867fce9182b6b36ac3b1926b254adoesn't exist in either net-next nor net trees.
Quoted 5 lines
Quoted 2 lines
My guess is you applied Maxime's patch locally, and that is the commit ID of that patch.
This was supposed to be the stable patch-id obtained with 'git patch-id --stable'.
Hmm, didn't know about that... but in this context, I wonder how useful it is. As a maintainer, given that patches submitted don't specify their patch-id, tracking down which patch is the pre-requisit would be a mammoth task, whereas a lore.kernel.org link would indicate the pre-requisit immediately.
lore.kernel.org links are easy:
https://lore.kernel.org/r/<message id of the email containing the patch>
I can see patch-ids are useful for automation, but lore.kernel.org URLs are useful for humans.
Signature · 2 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/ FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!
Jakub Kicinski
maintainerQuoted 7 lines · On Mon, 5 Jan 2026 17:13:45 +0000 Russell King (Oracle) wrote:
Quoted 7 lines
Quoted 6 lines
This seems to be almost useless information. While base-commit exists in the net-next tree, commit
ae039dad1e17867fce9182b6b36ac3b1926b254adoesn't exist in either net-next nor net trees.My guess is you applied Maxime's patch locally, and that is the commit ID of that patch.
Quoted 7 lines
Quoted 2 lines
This was supposed to be the stable patch-id obtained with 'git patch-id --stable'.
Hmm, didn't know about that... but in this context, I wonder how useful it is. As a maintainer, given that patches submitted don't specify their patch-id, tracking down which patch is the pre-requisit would be a mammoth task [...]
+1, please wait for any pre-requisites to be in the tree before posting. If you'd like to get early reviews send as RFC. Note that net gets merged into net-next every Thu (in case the dependency is cross-tree this is an extra wait).