| Message ID | 20220126220733.6356-1-bb@ti.com |
|---|---|
| State | Accepted |
| Delegated to: | Tom Rini |
| Headers | show |
| Series | soc: soc_ti_k3: update j721e revision numbering | expand |
Sorry Tom! On this day, January 26, 2022, thus sayeth Bryan Brattlof: > There is a 4 bit VARIANT number inside the JTAGID register that TI > increments any time a new variant for a chip is produced. Each > family of TI's SoCs uses a different versioning scheme based off > that VARIANT number. > > CC: Dave Gerlach <d-gerlach@ti.com> > Signed-off-by: Bryan Brattlof <bb@ti.com> > --- > drivers/soc/soc_ti_k3.c | 40 +++++++++++++++++++++++++++------------- > 1 file changed, 27 insertions(+), 13 deletions(-) > I should have sent this to you. ~Bryan
On Thu, Jan 27, 2022 at 01:01:33PM -0600, Bryan Brattlof wrote: > Sorry Tom! > > On this day, January 26, 2022, thus sayeth Bryan Brattlof: > > There is a 4 bit VARIANT number inside the JTAGID register that TI > > increments any time a new variant for a chip is produced. Each > > family of TI's SoCs uses a different versioning scheme based off > > that VARIANT number. > > > > CC: Dave Gerlach <d-gerlach@ti.com> > > Signed-off-by: Bryan Brattlof <bb@ti.com> > > --- > > drivers/soc/soc_ti_k3.c | 40 +++++++++++++++++++++++++++------------- > > 1 file changed, 27 insertions(+), 13 deletions(-) > > > > I should have sent this to you. I saw on the list, thanks.
On Wed, Jan 26, 2022 at 04:07:33PM -0600, Bryan Brattlof wrote: > There is a 4 bit VARIANT number inside the JTAGID register that TI > increments any time a new variant for a chip is produced. Each > family of TI's SoCs uses a different versioning scheme based off > that VARIANT number. > > CC: Dave Gerlach <d-gerlach@ti.com> > Signed-off-by: Bryan Brattlof <bb@ti.com> Applied to u-boot/master, thanks!
diff --git a/drivers/soc/soc_ti_k3.c b/drivers/soc/soc_ti_k3.c index 9abed7d490a2c..7a126857cd709 100644 --- a/drivers/soc/soc_ti_k3.c +++ b/drivers/soc/soc_ti_k3.c @@ -15,9 +15,6 @@ #define J7200 0xbb6d #define AM64X 0xbb38 -#define REV_SR1_0 0 -#define REV_SR2_0 1 - #define JTAG_ID_VARIANT_SHIFT 28 #define JTAG_ID_VARIANT_MASK (0xf << 28) #define JTAG_ID_PARTNO_SHIFT 12 @@ -55,25 +52,42 @@ static const char *get_family_string(u32 idreg) return family; } +static char *j721e_rev_string_map[] = { + "1.0", "1.1", +}; + +static char *am65x_rev_string_map[] = { + "1.0", "2.0", +}; + static const char *get_rev_string(u32 idreg) { - const char *revision; u32 rev; + u32 soc; rev = (idreg & JTAG_ID_VARIANT_MASK) >> JTAG_ID_VARIANT_SHIFT; + soc = (idreg & JTAG_ID_PARTNO_MASK) >> JTAG_ID_PARTNO_SHIFT; - switch (rev) { - case REV_SR1_0: - revision = "1.0"; - break; - case REV_SR2_0: - revision = "2.0"; - break; + switch (soc) { + case J721E: + if (rev > ARRAY_SIZE(j721e_rev_string_map)) + goto bail; + return j721e_rev_string_map[rev]; + + case AM65X: + if (rev > ARRAY_SIZE(am65x_rev_string_map)) + goto bail; + return am65x_rev_string_map[rev]; + + case AM64X: + case J7200: default: - revision = "Unknown Revision"; + if (!rev) + return "1.0"; }; - return revision; +bail: + return "Unknown Revision"; } static int soc_ti_k3_get_family(struct udevice *dev, char *buf, int size)
There is a 4 bit VARIANT number inside the JTAGID register that TI increments any time a new variant for a chip is produced. Each family of TI's SoCs uses a different versioning scheme based off that VARIANT number. CC: Dave Gerlach <d-gerlach@ti.com> Signed-off-by: Bryan Brattlof <bb@ti.com> --- drivers/soc/soc_ti_k3.c | 40 +++++++++++++++++++++++++++------------- 1 file changed, 27 insertions(+), 13 deletions(-)