software/src/opensbi.git/include, branch master

lib: sbi_domain: Rename per-domain data to per-domain state

2026-09-05T07:23:15+00:00

The per-domain sbi_domain_data hold each domain associated state like hart context in sbi_domain_context, mpxy state in sbi_mpxy, and others like each domain backed by the corresponding MPT (SMMPT). DATA reads as a generic name, while every use stores state. Rename functions and macros appropriately. There are no functional changes. Signed-off-by: Rahul Pathak Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260820121309.2551296-3-rahul.pathak@oss.qualcomm.com Signed-off-by: Anup Patel

lib: sbi: Rename map_range/unmap_range functions

2026-09-05T07:22:39+00:00

*_map_range and *_unmap_range functions are required in M-Mode to get the temporary access to S-Mode and U-Mode regions. Thse functions only operate using the TYPE_MEMORY memory protection mechanisms. Rename them to reflect the actual usage of these functions and let generic map/unmap_range names to implement generic functions Signed-off-by: Rahul Pathak Reviewed-by: Anup Patel Reviewed-by: Pawandeep Oza Link: https://lore.kernel.org/r/20260820121309.2551296-2-rahul.pathak@oss.qualcomm.com Signed-off-by: Anup Patel

lib: sbi_pmu: Fix counter and event info error codes as per SBI v3.0 spec

2026-09-04T06:13:53+00:00

Align the PMU extension implementation with the error codes required by the SBI v3.0 specification, chapter 11: - sbi_pmu_counter_start and sbi_pmu_counter_stop (secs 11.9-11.10, tables 39-42): the start_flags/stop_flags bits 2:(XLEN-1) are reserved and must be zero, so return SBI_ERR_INVALID_PARAM when any reserved flag bit is set. Introduce SBI_PMU_START_FLAGS_MASK and SBI_PMU_STOP_FLAGS_MASK for the valid bits of each function. - sbi_pmu_counter_start and sbi_pmu_counter_stop (tables 40 and 42): return SBI_ERR_ALREADY_STARTED / SBI_ERR_ALREADY_STOPPED when the set of counters includes a counter which is already started or stopped, instead of ignoring the error returned for each counter. - sbi_pmu_event_get_info (sec 11.14, table 47): the output word must indicate whether the event is supported, but firmware events were only matched against the hardware event map and were always reported as unsupported. Report a validated firmware event as supported. Signed-off-by: David E. Garcia Porras Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260818210023.466462-4-david.garcia@aheadcomputing.com Signed-off-by: Anup Patel

lib: sbi_pmu: Return invalid param error for reserved event_idx bits

2026-09-03T07:57:45+00:00

As per section 11.14 of the SBI specification (Function: Get PMU Event Info, FID #8), Table 47, the event_idx word of an event info entry only uses BIT[0:19]; BIT[20:31] are reserved for the future purpose and must be zero. Table 48 further requires the SBI implementation to return SBI_ERR_INVALID_PARAM if any reserved bit in an event_idx word is set. sbi_pmu_event_get_info() does not check the reserved bits, so a malformed event_idx is silently passed on to pmu_event_validate() instead of failing the call. Add SBI_PMU_EVENT_IDX_MBZ_MASK covering the must-be-zero bits and return SBI_ERR_INVALID_PARAM when any of them are set. Fixes: e4345842168b ("lib: sbi_pmu: Implement SBI PMU event info function") Signed-off-by: David E. Garcia Porras Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260818210023.466462-2-david.garcia@aheadcomputing.com Signed-off-by: Anup Patel

lib: utils/suspend: wait for the secondary Andes harts to sleep

2026-09-01T05:16:23+00:00

A secondary hart may not have reached the target sleep state by the time the primary checks, so the one-shot check could fail spuriously. Poll each PCS status instead and give up after HART_SLEEP_TIMEOUT_MS. Rework atcsmu_pcs_is_sleep() into the atcsmu_hart_is_sleep() predicate so sbi_timer_waitms_until() can drive it. Signed-off-by: Ben Zong-You Xie Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260728081041.2724668-7-ben717@andestech.com Signed-off-by: Anup Patel

lib: utils/hsm: wake sleeping Andes harts with an IPI

2026-09-01T05:16:23+00:00

An IPI wakes a hart from light sleep and, once MSIP is kept as an SMU wakeup event, from deep sleep too. Drop the WAKEUP_CMD path and the boot state, hart type and sleep type conditions that selected it. Signed-off-by: Ben Zong-You Xie Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260728081041.2724668-6-ben717@andestech.com Signed-off-by: Anup Patel

lib: sbi: dbtr: add platform device for per-slot trigger capabilities

2026-09-01T05:06:07+00:00

The tinfo probe only discovers trigger types; WARL fields tied off inside an implemented trigger CSR (e.g. an mcontrol6 trigger without load/store address match) are not discoverable. sbi_alloc_trigger() also selects the first free trigger slot irrespective of the requested trigger type or configuration. Unsupported configurations are thus silently dropped by the hardware while SBI reports success, where SBI v3.0 sections 19.4 / 19.5 require SBI_ERR_NOT_SUPPORTED. Add an optional platform device, struct sbi_dbtr_device, with a trigger_supported(idx, tdata1, tdata2, tdata3) callback reporting whether the trigger slot selected by idx (the tselect value) supports a given configuration. Use it, together with the probed per-slot type_mask, to: - allocate only trigger slots supporting the requested configuration (new dbtr_find_free_slot() helper) - reject unsupportable install/update requests with SBI_ERR_NOT_SUPPORTED - count only supporting slots in sbi_dbtr_num_trig() (SBI v3.0 section 19.1) The install path now dry-runs the allocation of the whole batch before programming any trigger, so a partially-installed batch is never left behind. Without a registered device, only the type matching applies (unchanged behavior). Fixes: 97f234f15c96 ("lib: sbi: Introduce the SBI debug triggers extension support") Signed-off-by: David E. Garcia Porras Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260727183041.258377-2-david.garcia@aheadcomputing.com Signed-off-by: Anup Patel

lib: sbi: Allow platforms to override local TLB operations

2026-08-20T15:00:48+00:00

Some T-Head based processors, for example the Sophgo SG2044, have a JTLB errata that requires special handling of certain TLB maintenance instructions (sfence.vma). Introduce a sbi_tlb_local_operations structure so platforms can provide custom local TLB flush implementations. Use it to implement the JTLB workaround on affected SoCs (currently Sophgo SG2044). Signed-off-by: Xiang W Signed-off-by: Han Gao Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260617112925.1144190-2-wangxiang@iscas.ac.cn Signed-off-by: Anup Patel

lib/sbi/sbi_timer: Add sbi_timer_frequency() function

2026-08-20T12:27:45+00:00

Drivers and other library code sometimes need the timer frequency outside of lib/sbi/sbi_timer.c, but there is currently no accessor for it. Add sbi_timer_frequency() to expose the frequency of the registered timer device. Signed-off-by: Sunil V L Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260703050530.2460037-2-sunilvl@oss.qualcomm.com Signed-off-by: Anup Patel

lib: make SBI_HARTMASK_MAX_BITS customizable, default 256

2026-08-20T12:27:19+00:00

Linux has raised the default number of 64bit CPUs to 256 in commit 0fa749771993 ("riscv: Raise default NR_CPUS for 64BIT to 256"). OpenSBI should use a matching value for SBI_HARTMASK_MAX_BITS. Introduce configuration variable CONFIG_HARTMASK_MAX_BITS with range 32-512 and default 256. Signed-off-by: Heinrich Schuchardt Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260706160544.697409-1-heinrich.schuchardt@canonical.com Signed-off-by: Anup Patel