| Message ID | 20241023055118.1400286-6-Shyam-sundar.S-k@amd.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | Introduce initial support for the AMD I3C (non-HCI) to DW driver | expand |
On 10/23/24 8:51 AM, Shyam Sundar S K wrote: > I3C devices like DIMMs over SPD use SETAASA for bus discovery instead of > SETDASA. Add a new routine for I3C host controller drivers to use. If the > I3C slave on the bus is an SPD device, skip the regular DAA process. > > According to the SPD spec[1], use SETAASA for bus discovery, and avoid > sending RSTDAA and DISEC, as they are considered illegal. Skip this entire > process if the slave is SPD-compliant, as indicated by the "jdec_spd" flag > from the BIOS. > > [1] https://www.jedec.org/system/files/docs/JESD300-5B.01.pdf > (section 2.4 and 2.6.3) > SETAASA seems to come in MIPI v1.1.1 specification. I think worth to mention in commit log. > Co-developed-by: Sanket Goswami <Sanket.Goswami@amd.com> > Signed-off-by: Sanket Goswami <Sanket.Goswami@amd.com> > Signed-off-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> > --- > drivers/i3c/master.c | 32 +++++++++++++++++++++++++++++- > drivers/i3c/master/dw-i3c-master.c | 1 + > include/linux/i3c/ccc.h | 1 + > include/linux/i3c/master.h | 1 + > 4 files changed, 34 insertions(+), 1 deletion(-) > > diff --git a/drivers/i3c/master.c b/drivers/i3c/master.c > index ba6f17cb8aa6..1596efd6d82a 100644 > --- a/drivers/i3c/master.c > +++ b/drivers/i3c/master.c > @@ -1657,6 +1657,21 @@ i3c_master_register_new_i3c_devs(struct i3c_master_controller *master) > } > } > > +static int i3c_master_setaasa_locked(struct i3c_master_controller *master) > +{ > + struct i3c_ccc_cmd_dest dest; > + struct i3c_ccc_cmd cmd; > + int ret; > + > + i3c_ccc_cmd_dest_init(&dest, I3C_BROADCAST_ADDR, 0); > + i3c_ccc_cmd_init(&cmd, false, I3C_CCC_SETAASA, &dest, 1); > + > + ret = i3c_master_send_ccc_cmd_locked(master, &cmd); > + i3c_ccc_cmd_dest_cleanup(&dest); > + > + return ret; > +} > + > static int i3c_master_add_spd_dev(struct i3c_master_controller *master, > struct i3c_dev_boardinfo *boardinfo) > { > @@ -1684,6 +1699,10 @@ static int i3c_master_add_spd_dev(struct i3c_master_controller *master, > i3cdev->info.pid = i3cdev->boardinfo->pid; > i3cdev->info.dyn_addr = i3cdev->boardinfo->init_dyn_addr; > > + ret = i3c_master_setaasa_locked(master); > + if (ret) > + goto err_free_dev; > + > i3c_bus_normaluse_lock(&master->bus); > i3c_master_register_new_i3c_devs(master); > i3c_bus_normaluse_unlock(&master->bus); > @@ -1907,7 +1926,14 @@ static int i3c_master_bus_init(struct i3c_master_controller *master) > goto err_bus_cleanup; > } > > - i3c_master_add_spd_dev(master, i3cboardinfo); > + /* > + * If the I3C slave on the bus is SPD device, then do not follow the regular > + * DAA process. Also, as per SPD spec SETAASA is required for the bus discovery > + * and sending RSTDAA and DISEC is considered as illegal. So skip the entire process > + * if the jdec_spd flag has been identified from the BIOS. > + */ > + if (master->jdec_spd) > + return i3c_master_add_spd_dev(master, i3cboardinfo); > > if (master->ops->set_speed) { > ret = master->ops->set_speed(master, I3C_OPEN_DRAIN_SLOW_SPEED); > @@ -2311,6 +2337,10 @@ static int i3c_acpi_configure_master(struct i3c_master_controller *master) > return -ENODEV; > } > > + status = acpi_evaluate_object(master->ahandle, "_STR", NULL, NULL); > + if (ACPI_SUCCESS(status)) > + master->jdec_spd = true; > + Am I right "_STR" object should carry a string "jdec_spd"? But this code is not actually checking it, only the existence of _STR? > num_dev = device_get_child_node_count(dev); > if (!num_dev) { > dev_err(&master->dev, "Error: no child node present\n"); > diff --git a/drivers/i3c/master/dw-i3c-master.c b/drivers/i3c/master/dw-i3c-master.c > index f683e2a398ad..90a43209e55e 100644 > --- a/drivers/i3c/master/dw-i3c-master.c > +++ b/drivers/i3c/master/dw-i3c-master.c > @@ -282,6 +282,7 @@ static bool dw_i3c_master_supports_ccc_cmd(struct i3c_master_controller *m, > case I3C_CCC_GETSTATUS: > case I3C_CCC_GETMXDS: > case I3C_CCC_GETHDRCAP: > + case I3C_CCC_SETAASA: > return true; > default: > return false; > diff --git a/include/linux/i3c/ccc.h b/include/linux/i3c/ccc.h > index ad59a4ae60d1..a145d766ab6f 100644 > --- a/include/linux/i3c/ccc.h > +++ b/include/linux/i3c/ccc.h > @@ -32,6 +32,7 @@ > #define I3C_CCC_DEFSLVS I3C_CCC_ID(0x8, true) > #define I3C_CCC_ENTTM I3C_CCC_ID(0xb, true) > #define I3C_CCC_ENTHDR(x) I3C_CCC_ID(0x20 + (x), true) > +#define I3C_CCC_SETAASA I3C_CCC_ID(0x29, true) > > /* Unicast-only commands */ > #define I3C_CCC_SETDASA I3C_CCC_ID(0x7, false) > diff --git a/include/linux/i3c/master.h b/include/linux/i3c/master.h > index 367faf7c4bf3..cd8390d8b469 100644 > --- a/include/linux/i3c/master.h > +++ b/include/linux/i3c/master.h > @@ -516,6 +516,7 @@ struct i3c_master_controller { > const struct i3c_master_controller_ops *ops; > unsigned int secondary : 1; > unsigned int init_done : 1; > + unsigned int jdec_spd : 1; > unsigned int hotjoin: 1; > struct { > struct list_head i3c;
On 11/4/2024 20:30, Jarkko Nikula wrote: > On 10/23/24 8:51 AM, Shyam Sundar S K wrote: >> I3C devices like DIMMs over SPD use SETAASA for bus discovery >> instead of >> SETDASA. Add a new routine for I3C host controller drivers to use. >> If the >> I3C slave on the bus is an SPD device, skip the regular DAA process. >> >> According to the SPD spec[1], use SETAASA for bus discovery, and avoid >> sending RSTDAA and DISEC, as they are considered illegal. Skip this >> entire >> process if the slave is SPD-compliant, as indicated by the >> "jdec_spd" flag >> from the BIOS. >> >> [1] https://www.jedec.org/system/files/docs/JESD300-5B.01.pdf >> (section 2.4 and 2.6.3) >> > SETAASA seems to come in MIPI v1.1.1 specification. I think worth to > mention in commit log. > OK. I will do it in the next version. >> Co-developed-by: Sanket Goswami <Sanket.Goswami@amd.com> >> Signed-off-by: Sanket Goswami <Sanket.Goswami@amd.com> >> Signed-off-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> >> --- >> drivers/i3c/master.c | 32 >> +++++++++++++++++++++++++++++- >> drivers/i3c/master/dw-i3c-master.c | 1 + >> include/linux/i3c/ccc.h | 1 + >> include/linux/i3c/master.h | 1 + >> 4 files changed, 34 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/i3c/master.c b/drivers/i3c/master.c >> index ba6f17cb8aa6..1596efd6d82a 100644 >> --- a/drivers/i3c/master.c >> +++ b/drivers/i3c/master.c >> @@ -1657,6 +1657,21 @@ i3c_master_register_new_i3c_devs(struct >> i3c_master_controller *master) >> } >> } >> +static int i3c_master_setaasa_locked(struct i3c_master_controller >> *master) >> +{ >> + struct i3c_ccc_cmd_dest dest; >> + struct i3c_ccc_cmd cmd; >> + int ret; >> + >> + i3c_ccc_cmd_dest_init(&dest, I3C_BROADCAST_ADDR, 0); >> + i3c_ccc_cmd_init(&cmd, false, I3C_CCC_SETAASA, &dest, 1); >> + >> + ret = i3c_master_send_ccc_cmd_locked(master, &cmd); >> + i3c_ccc_cmd_dest_cleanup(&dest); >> + >> + return ret; >> +} >> + >> static int i3c_master_add_spd_dev(struct i3c_master_controller >> *master, >> struct i3c_dev_boardinfo *boardinfo) >> { >> @@ -1684,6 +1699,10 @@ static int i3c_master_add_spd_dev(struct >> i3c_master_controller *master, >> i3cdev->info.pid = i3cdev->boardinfo->pid; >> i3cdev->info.dyn_addr = i3cdev->boardinfo->init_dyn_addr; >> + ret = i3c_master_setaasa_locked(master); >> + if (ret) >> + goto err_free_dev; >> + >> i3c_bus_normaluse_lock(&master->bus); >> i3c_master_register_new_i3c_devs(master); >> i3c_bus_normaluse_unlock(&master->bus); >> @@ -1907,7 +1926,14 @@ static int i3c_master_bus_init(struct >> i3c_master_controller *master) >> goto err_bus_cleanup; >> } >> - i3c_master_add_spd_dev(master, i3cboardinfo); >> + /* >> + * If the I3C slave on the bus is SPD device, then do not >> follow the regular >> + * DAA process. Also, as per SPD spec SETAASA is required for >> the bus discovery >> + * and sending RSTDAA and DISEC is considered as illegal. So >> skip the entire process >> + * if the jdec_spd flag has been identified from the BIOS. >> + */ >> + if (master->jdec_spd) >> + return i3c_master_add_spd_dev(master, i3cboardinfo); >> if (master->ops->set_speed) { >> ret = master->ops->set_speed(master, >> I3C_OPEN_DRAIN_SLOW_SPEED); >> @@ -2311,6 +2337,10 @@ static int i3c_acpi_configure_master(struct >> i3c_master_controller *master) >> return -ENODEV; >> } >> + status = acpi_evaluate_object(master->ahandle, "_STR", NULL, >> NULL); >> + if (ACPI_SUCCESS(status)) >> + master->jdec_spd = true; >> + > > Am I right "_STR" object should carry a string "jdec_spd"? But this > code is not actually checking it, only the existence of _STR? > I don't think so.. its just the string SPD/DIMM. Scope (\_SB.I3CA) { Name (_STR, Unicode ("SPD/DIMM")) // _STR: Description String Name (_DSD, Package (0x04) // _DSD: Device-Specific Data { ... }) Name (MST0, Package (0x02) { ... }) Device (SPD0) { ... } Device (PMI0) { ... } Device (RCD0) { ... } } Thanks, Shyam >> num_dev = device_get_child_node_count(dev); >> if (!num_dev) { >> dev_err(&master->dev, "Error: no child node present\n"); >> diff --git a/drivers/i3c/master/dw-i3c-master.c >> b/drivers/i3c/master/dw-i3c-master.c >> index f683e2a398ad..90a43209e55e 100644 >> --- a/drivers/i3c/master/dw-i3c-master.c >> +++ b/drivers/i3c/master/dw-i3c-master.c >> @@ -282,6 +282,7 @@ static bool >> dw_i3c_master_supports_ccc_cmd(struct i3c_master_controller *m, >> case I3C_CCC_GETSTATUS: >> case I3C_CCC_GETMXDS: >> case I3C_CCC_GETHDRCAP: >> + case I3C_CCC_SETAASA: >> return true; >> default: >> return false; >> diff --git a/include/linux/i3c/ccc.h b/include/linux/i3c/ccc.h >> index ad59a4ae60d1..a145d766ab6f 100644 >> --- a/include/linux/i3c/ccc.h >> +++ b/include/linux/i3c/ccc.h >> @@ -32,6 +32,7 @@ >> #define I3C_CCC_DEFSLVS I3C_CCC_ID(0x8, true) >> #define I3C_CCC_ENTTM I3C_CCC_ID(0xb, true) >> #define I3C_CCC_ENTHDR(x) I3C_CCC_ID(0x20 + (x), true) >> +#define I3C_CCC_SETAASA I3C_CCC_ID(0x29, true) >> /* Unicast-only commands */ >> #define I3C_CCC_SETDASA I3C_CCC_ID(0x7, false) >> diff --git a/include/linux/i3c/master.h b/include/linux/i3c/master.h >> index 367faf7c4bf3..cd8390d8b469 100644 >> --- a/include/linux/i3c/master.h >> +++ b/include/linux/i3c/master.h >> @@ -516,6 +516,7 @@ struct i3c_master_controller { >> const struct i3c_master_controller_ops *ops; >> unsigned int secondary : 1; >> unsigned int init_done : 1; >> + unsigned int jdec_spd : 1; >> unsigned int hotjoin: 1; >> struct { >> struct list_head i3c; >
diff --git a/drivers/i3c/master.c b/drivers/i3c/master.c index ba6f17cb8aa6..1596efd6d82a 100644 --- a/drivers/i3c/master.c +++ b/drivers/i3c/master.c @@ -1657,6 +1657,21 @@ i3c_master_register_new_i3c_devs(struct i3c_master_controller *master) } } +static int i3c_master_setaasa_locked(struct i3c_master_controller *master) +{ + struct i3c_ccc_cmd_dest dest; + struct i3c_ccc_cmd cmd; + int ret; + + i3c_ccc_cmd_dest_init(&dest, I3C_BROADCAST_ADDR, 0); + i3c_ccc_cmd_init(&cmd, false, I3C_CCC_SETAASA, &dest, 1); + + ret = i3c_master_send_ccc_cmd_locked(master, &cmd); + i3c_ccc_cmd_dest_cleanup(&dest); + + return ret; +} + static int i3c_master_add_spd_dev(struct i3c_master_controller *master, struct i3c_dev_boardinfo *boardinfo) { @@ -1684,6 +1699,10 @@ static int i3c_master_add_spd_dev(struct i3c_master_controller *master, i3cdev->info.pid = i3cdev->boardinfo->pid; i3cdev->info.dyn_addr = i3cdev->boardinfo->init_dyn_addr; + ret = i3c_master_setaasa_locked(master); + if (ret) + goto err_free_dev; + i3c_bus_normaluse_lock(&master->bus); i3c_master_register_new_i3c_devs(master); i3c_bus_normaluse_unlock(&master->bus); @@ -1907,7 +1926,14 @@ static int i3c_master_bus_init(struct i3c_master_controller *master) goto err_bus_cleanup; } - i3c_master_add_spd_dev(master, i3cboardinfo); + /* + * If the I3C slave on the bus is SPD device, then do not follow the regular + * DAA process. Also, as per SPD spec SETAASA is required for the bus discovery + * and sending RSTDAA and DISEC is considered as illegal. So skip the entire process + * if the jdec_spd flag has been identified from the BIOS. + */ + if (master->jdec_spd) + return i3c_master_add_spd_dev(master, i3cboardinfo); if (master->ops->set_speed) { ret = master->ops->set_speed(master, I3C_OPEN_DRAIN_SLOW_SPEED); @@ -2311,6 +2337,10 @@ static int i3c_acpi_configure_master(struct i3c_master_controller *master) return -ENODEV; } + status = acpi_evaluate_object(master->ahandle, "_STR", NULL, NULL); + if (ACPI_SUCCESS(status)) + master->jdec_spd = true; + num_dev = device_get_child_node_count(dev); if (!num_dev) { dev_err(&master->dev, "Error: no child node present\n"); diff --git a/drivers/i3c/master/dw-i3c-master.c b/drivers/i3c/master/dw-i3c-master.c index f683e2a398ad..90a43209e55e 100644 --- a/drivers/i3c/master/dw-i3c-master.c +++ b/drivers/i3c/master/dw-i3c-master.c @@ -282,6 +282,7 @@ static bool dw_i3c_master_supports_ccc_cmd(struct i3c_master_controller *m, case I3C_CCC_GETSTATUS: case I3C_CCC_GETMXDS: case I3C_CCC_GETHDRCAP: + case I3C_CCC_SETAASA: return true; default: return false; diff --git a/include/linux/i3c/ccc.h b/include/linux/i3c/ccc.h index ad59a4ae60d1..a145d766ab6f 100644 --- a/include/linux/i3c/ccc.h +++ b/include/linux/i3c/ccc.h @@ -32,6 +32,7 @@ #define I3C_CCC_DEFSLVS I3C_CCC_ID(0x8, true) #define I3C_CCC_ENTTM I3C_CCC_ID(0xb, true) #define I3C_CCC_ENTHDR(x) I3C_CCC_ID(0x20 + (x), true) +#define I3C_CCC_SETAASA I3C_CCC_ID(0x29, true) /* Unicast-only commands */ #define I3C_CCC_SETDASA I3C_CCC_ID(0x7, false) diff --git a/include/linux/i3c/master.h b/include/linux/i3c/master.h index 367faf7c4bf3..cd8390d8b469 100644 --- a/include/linux/i3c/master.h +++ b/include/linux/i3c/master.h @@ -516,6 +516,7 @@ struct i3c_master_controller { const struct i3c_master_controller_ops *ops; unsigned int secondary : 1; unsigned int init_done : 1; + unsigned int jdec_spd : 1; unsigned int hotjoin: 1; struct { struct list_head i3c;