common: cmd_dump: Gracefully handle ti_sci_cmd_get_clk_state() failure If one has a kernel with configs like this: CONFIG_DEVMEM=y CONFIG_GENERIC_LIB_DEVMEM_IS_ALLOWED=y CONFIG_STRICT_DEVMEM=y CONFIG_IO_STRICT_DEVMEM=y then mmap'ing /dev/mem will fail. This results in all sort of issues, but eventually will lead to a segfault if invoked like so on the j784s4-evm: [root@fedora build]# k3conf dump clock 218 Map fail Map fail Map fail k3_sec_proxy_send: Thread23 verification failed. ret = -2 |-----------------------------------------------------------------------| | VERSION INFO | |-----------------------------------------------------------------------| | K3CONF | (version v0.3-4-g16ee9a4 built Tue Mar 05 15:48:09 UTC 2024) | | SoC | J784S4 SR1.0 | |-----------------------------------------------------------------------| Map fail Map fail Map fail k3_sec_proxy_send: Thread23 verification failed. ret = -2 Segmentation fault (core dumped) [root@fedora build]# Let's at least avoid the segfault by not using the NULL pointer returned by ti_sci_cmd_get_clk_state(). Signed-off-by: Andrew Halaney <ahalaney@redhat.com>
readme: Add disclaimer about /dev/mem usage Users may want to know up front that this tool needs access to /dev/mem, in particular to talk over the mailboxes. Without this users may face errors. This includes making sure CONFIG_DEVMEM_STRICT_IO is disabled, as most systems will actively be using the mailboxes necessary to communicate. Signed-off-by: Andrew Halaney <ahalaney@redhat.com>
soc: Add support for J722S Add SoC data and initial support for J722S device family. Signed-off-by: Vaishnav Achath <vaishnav.a@ti.com>
common: Fix 32-bit print & scan issues The code doesn't work on 32-bit builds as the printf & scanf formats used are not correct. Fix these by using the formatting macros. Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Reviewed-by: Andrew Davis <afd@ti.com>
CMakeLists.txt: Bump to 0.3 Major changes in the release: - New supported SoCs - Am62x, AM62Ax, AM62Px - J721S2/AM68, J784S4/AM69 - New features added - Switch over to cmake - (experimental) DDRBW for AM62x, J721E, J7200 - Set clock parent option - Add a proper LICENSE file - Upgrades and Bug fixes: - Tons of refactor of code - Fixes across processor name, register read/write for 32bit regs - Bunch of cosmetic fixups Signed-off-by: Nishanth Menon <nm@ti.com>
soc: am62px: add secure proxy descriptions Provide the secure proxy and communication path descriptions that are allowed for the AM62Px SoC family Signed-off-by: Bryan Brattlof <bb@ti.com>
soc: am62px: add resource assignment descriptions The resource type IDs represent the resource ranges that are assignable to the AM62Px's processing entities. Provide the board configuration resource assignment type IDs that are permitted in the AM62Px SoC family Signed-off-by: Bryan Brattlof <bb@ti.com>
soc: am62px: add processors information The Processor IDs (not to be confused with Host IDs) represent the *actual* physical processors in the AM62Px Provide the Processor IDs available in the AM62Px SoC family Signed-off-by: Bryan Brattlof <bb@ti.com>
soc: am62px: add host description information These Host IDs represent processing entities for the AM62Px. Typically a host is defined as a 'processing entity' which may be an actual processor or could represent a virtual machine. Provide the Host IDs permitted in the AM62Px SoC family Signed-off-by: Bryan Brattlof <bb@ti.com>
soc: am62px: add device description information The Device IDs represent the SoC subsystems that can be modified by the TISCI message API. Some subsystems also use a Device ID as a parameter, allowing us to specify the management of a particular SoC subsystem. Provide the Device IDs permitted in the AM62Px SoC family. Signed-off-by: Bryan Brattlof <bb@ti.com>
soc: am62px: add clock identifiers information The TISCI power management API uses Device IDs and Clock IDs as parameters, allowing us granular control of the clocks for a particular subsystem. Define the Clock IDs that identify the incoming and outgoing clock signals for subsystems in the AM62Px family of SoCs Signed-off-by: Bryan Brattlof <bb@ti.com>
common: help: Remove leading newline and tabs escape sequences These can make it harder to see what the output will look like from only looking at the code. They also confuse tools like codespell. Move leading newlines to a new line. Replace tab escape sequences with real tabs. Should result in no functional change to the help output. Signed-off-by: Andrew Davis <afd@ti.com>