Because *out needs to reserve a byte to hold '\0', no more characters
should be added to the buffer when *out has one byte left, and the
buffer size *out_len should not be modified. this patch prevents
the correction of *out_len when *out_len is 1.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
When doing width = width - strlen(string) in prints there is no need
to consider the case that witdh may be less than 0. This is because
the code to do filling needs to be executed under the condition that
width > 0.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The information of sg/b/letbase can be obtained by the type character,
simplifying the parameter by passing the type directly.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The space flag is used to add a space before positive numbers, and
apostrophe is used to print the thousand separator. Add code to
ignore these two flags
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Adds + flags for print, prefixing positive numbers with + when this
flags is present
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Left alignment and padding '0' should not exist at the same time,
this patch skips padding.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The flags for print should be able to appear in any order. The
previous code required the order to be fixed.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Fix Priv spec version typo in commit d4b563c881 ("lib: sbi: Remove MCOUNTEREN
and SCOUNTEREN hart features").
At least Priv spec v1.11 is required for [m|s]counteren and mcountinhibit CSRs.
Fixes: d4b563c881 ("lib: sbi: Remove MCOUNTEREN and SCOUNTEREN hart features")
Signed-off-by: Ley Foon Tan <leyfoon.tan@starfivetech.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The hart_pmu_get_allowed_bits() function detects implemented bits
of mhpm counters so let us rename this function accordingly.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Xiang W <wxjstz@126.com>
Let us follow alphabetical order for HART ISA extension so that
it is simpler to maintain.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Xiang W <wxjstz@126.com>
Recently ratified Zihpm ISA extension covers all [m]hpm* CSRs
so we add Zihpm as a HART ISA extension in OpenSBI.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Xiang W <wxjstz@126.com>
Recently ratified Zicntr ISA extension covers cycle, time and
instret CSRs so we replace the "time" ISA extension with "zicntr"
ISA extension in OpenSBI.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Xiang W <wxjstz@126.com>
On a misconfigured system we could access phs->active_events[] out of
bounds. Check that num_hw_ctrs is less or equal SBI_PMU_HW_CTR_MAX.
Addresses-Coverity-ID: 1566113 ("Out-of-bounds read")
Addresses-Coverity-ID: 1566114 ("Out-of-bounds write")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Do not ignore register A2 (high bits of physical address) in the dbcn
handler (RV64).
Signed-off-by: Gianluca Guida <gianluca@rivosinc.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Instead of using a global array indexed by hartid, we should use
scratch space to save per-HART domain pointer.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Let's use heap allocation in root domain creation instead of using
a fixed size global array.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Instead of using a global array for per-HART PMU state, we should
use heap to on-demand allocate per-HART PMU state when the HART
is initialized in cold boot or warm boot path.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
The scratch space being a scarce resource so let us print it's
size and usage at boot time.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
We provide simple heap allocator to manage the heap space provided
by OpenSBI firmware and platform.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
The spec says sbi_system_suspend() will return SBI_ERR_INVALID_PARAM
when "sleep_type is reserved or is platform-specific and unimplemented"
and SBI_ERR_NOT_SUPPORTED when sleep_type "is not reserved and is
implemented, but the platform does not support it due to one or more
missing dependencies." Ensure SBI_ERR_INVALID_PARAM is returned for
reserved sleep types and that the system suspend driver can choose
which of the two error types to return itself by returning an error
from its check function rather than a boolean.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
console is not a required peripheral. So it should return success when
the console does not exist.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
When a probe implementation just returns zero for not available and
one for available then we don't need it, as the extension won't be
registered at all if it would return zero and the Base extension
probe function will already set out_val to 1 if not probe function
is implemented. Currently all probe functions only return zero or
one, so remove them all.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
No need to do a fully comprehensive count, just find a supported reset
or suspend type
Signed-off-by: Xiang W <wxjstz@126.com>
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
When an extension implements a probe function it means there's a
chance that the extension is not available. Use this function in the
register_extensions callback to determine if the extension should be
registered at all. Where the probe implementation is simple, just
open code the check.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The absence of a probe implementation means that the extension is
always available. Remove the implementation for the PMU extension,
which does no checking, and indeed even has a comment saying it's
always available.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
The vendor extension ID range is large, but at runtime at most
a single ID will be available. Narrow the range in the
register_extensions callback. After narrowing, we no longer
need to check that the extension ID is correct in the other
callbacks, as those callbacks will never be invoked with
anything other than the single ID.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Rather than registering all extensions on their behalf in
sbi_ecall_init(), introduce another extension callback and
invoke that instead. For now, implement each callback by
simply registering the extension, which means this patch
has no intended functional change. In later patches, extension
callbacks will be modified to choose when to register and to
possibly narrow the extension ID range prior to registering.
When an extension range needs to remove IDs, leaving gaps, then
multiple invocations of sbi_ecall_register_extension() may be
used. In summary, later patches for current extensions and the
introductions of future extensions will use the new callback to
ensure that only valid extension IDs from the initial range,
which are also available, will be registered.
Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
We have redundant semicolon at quite a few places so let's remove it.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Since sbi_pmu_init is called after sbi_console_init,
the sbi_printf can be called when sbi_pmu_init fails.
Signed-off-by: Tan En De <ende.tan@starfivetech.com>
Reviewed-by: Atish Patra <atishp@rivosinc.com>
The legacy console getchar SBI call returns character value in
the sbiret.error field so the "SBI_SUCCESS < ret" check in
sbi_ecall_handler() results in unwanted error prints for the
legacy console getchar SBI call. Let's suppress these unwanted
error prints.
Fixes: 67b2a40892 ("lib: sbi: sbi_ecall: Check the range of
SBI error")
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
As per the SBI specification, we should "unconditionally select the first
counter from the set of counters specified by the counter_idx_base and
counter_idx_mask", so implement this behaviour.
Suggested-by: Atish Patra <atishp@atishpatra.org>
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Atish Patra <atishp@rivosinc.com>
Those events are enabled by default and should not be reset afterwards
since when using SBI_PMU_CFG_FLAG_SKIP_MATCH, it leads to unaccessible
counters after the first use.
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Atish Patra <atishp@rivosinc.com>
When tlb_fifo is full, it will wait and affect the ipi update to
other harts. This patch is optimized.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Tested-by: Anup Patel <anup@brainfault.org>
tlb_process_count is only used when count=1, so refactor to
tlb_process_once and add the return value to be reused in
tlb_process
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Tested-by: Anup Patel <anup@brainfault.org>
The original sbi_ipi will be processed by hart by hart, after optimization,
send ipi first and finally wait together.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Tested-by: Anup Patel <anup@brainfault.org>
Originally, the process and sync of sbi_tlb need to wait for each other.
Evasion by atomic addition and subtraction.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Tested-by: Anup Patel <anup@brainfault.org>
The bits encoded in event_idx[19:16] indicate the event type, with
an offset of 16 instead of 20.
Fixes: 13d40f21d5 ("lib: sbi: Add PMU support")
Signed-off-by: Yu Chien Peter Lin <peterlin@andestech.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Xiang W <wxjstz@126.com>
Implement SBI CPPC extension. This extension is only available when
OpenSBI platform provides a CPPC device to generic library.
Signed-off-by: Sunil V L <sunilvl@ventanamicro.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
If mip.SEIP bit is not cleared then on HiFive Unmatched board it causes
spurious external interrupts. This breaks the boot up of HiFive Unmatched
board. Hence it is required to bring the mip CSR to a known state during
hart init and avoid spurious interrupts.
Fixes: d9e7368 ("firmware: Not to clear all the MIP")
Signed-off-by: Mayuresh Chitale <mchitale@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
It possible that a platform supports hart hotplug (i.e. both hart_start
and hart_stop callbacks available) and all harts are start simultaneously
at platform boot-time. In this situation, the sbi_hsm_hart_start() will
call hsm_device_hart_start() for secondary harts at platform boot-time
which will fail because secondary harts were already started.
To fix above, we call hsm_device_hart_start() from sbi_hsm_hart_start()
only when entry_count is same as init_count for the secondary hart.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
We introduce sbi_entry_count() function which counts the number
of times a HART enters OpenSBI via cold-boot or warm-boot path.
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Because firmware is split into rw/rx segments, it cannot be recorded
by a root_fw_region. This problem is solved by adding a flag
fw_region_inited to sbi_domain.
Signed-off-by: Xiang W <wxjstz@126.com>
Reviewed-by: Himanshu Chauhan <hchauhan@ventanamicro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
Since commit 50d4fde1c5 ("lib: Remove redundant sbi_platform_ipi_clear()
calls"), the IPI sent from the boot hart in wake_coldboot_harts() is not
cleared in the secondary harts until they reach sbi_ipi_init(). However,
sbi_hsm_init() and sbi_hsm_hart_wait() are called earlier, so a secondary
hart might enter sbi_hsm_hart_wait() with an already pending IPI.
sbi_hsm_hart_wait() makes sure the hart leaves the loop only when it is
actually ready, so a pending unrelated IPI should not cause safety issues.
However, it might be inefficient on certain hardware, because it prevents
"wfi" from stalling the hart even if the hardware supports this, making the
hart needlessly spin in a "busy-wait" loop.
This behaviour can be observed, for example, in a QEMU VM (QEMU 7.2.0) with
"-machine virt" running a Linux guest. Inserting delays in
sbi_hsm_hart_start() allows reproducing the issue more reliably.
The comment in wait_for_coldboot() suggests that the initial IPI is needed
in the warm resume path, so let us clear it before init_warm_startup()
only.
To do this, sbi_ipi_raw_clear() was created similar to sbi_ipi_raw_send().
Signed-off-by: Evgenii Shatokhin <e.shatokhin@yadro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
When a boot hart executes sbi_hsm_hart_start() to start a secondary hart,
next_arg1, next_addr and next_mode for the latter are stored in the scratch
area after the state has been set to SBI_HSM_STATE_START_PENDING.
The secondary hart waits in the loop with wfi() in sbi_hsm_hart_wait() at
that time. However, "wfi" instruction is not guaranteed to wait for an
interrupt to be received by the hart, it is just a hint for the CPU.
According to RISC-V Privileged Architectures spec. v20211203, even an
implementation of "wfi" as "nop" is legal.
So, the secondary might leave the loop in sbi_hsm_hart_wait() as soon as
its state has been set to SBI_HSM_STATE_START_PENDING, even if it got no
IPI or it got an IPI unrelated to sbi_hsm_hart_start(). This could lead to
the following race condition when booting Linux, for example:
Boot hart (#0) Secondary hart (#1)
runs Linux startup code waits in sbi_hsm_hart_wait()
sbi_ecall(SBI_EXT_HSM,
SBI_EXT_HSM_HART_START,
...)
enters sbi_hsm_hart_start()
sets state of hart #1 to START_PENDING
leaves sbi_hsm_hart_wait()
runs to the end of init_warmboot()
returns to scratch->next_addr
(next_addr can be garbage here)
sets next_addr, etc. for hart #1
(no good: hart #1 has already left)
sends IPI to hart #1
(no good either)
If this happens, the secondary hart jumps to a wrong next_addr at the end
of init_warmboot(), which leads to a system hang or crash.
To reproduce the issue more reliably, one could add a delay in
sbi_hsm_hart_start() after setting the hart's state but before sending
IPI to that hart:
hstate = atomic_cmpxchg(&hdata->state, SBI_HSM_STATE_STOPPED,
SBI_HSM_STATE_START_PENDING);
...
+ sbi_timer_mdelay(10);
init_count = sbi_init_count(hartid);
rscratch->next_arg1 = arg1;
rscratch->next_addr = saddr;
The issue can be reproduced, for example, in a QEMU VM with '-machine virt'
and 2 or more CPUs, with Linux as the guest OS.
This patch moves writing of next_arg1, next_addr and next_mode for the
secondary hart before setting its state to SBI_HSM_STATE_START_PENDING.
In theory, it is possible that two or more harts enter sbi_hsm_hart_start()
for the same target hart simultaneously. To make sure the current hart has
exclusive access to the scratch area of the target hart at that point, a
per-hart 'start_ticket' is used. It is initially 0. The current hart tries
to acquire the ticket first (set it to 1) at the beginning of
sbi_hsm_hart_start() and only proceeds if it has successfully acquired it.
The target hart reads next_addr, etc., and then the releases the ticket
(sets it to 0) before calling sbi_hart_switch_mode(). This way, even if
some other hart manages to enter sbi_hsm_hart_start() after the ticket has
been released but before the target hart jumps to next_addr, it will not
cause problems.
atomic_cmpxchg() already has "acquire" semantics, among other things, so
no additional barriers are needed in hsm_start_ticket_acquire(). No hart
can perform or observe the update of *rscratch before setting of
'start_ticket' to 1.
atomic_write() only imposes ordering of writes, so an explicit barrier is
needed in hsm_start_ticket_release() to ensure its "release" semantics.
This guarantees that reads of scratch->next_addr, etc., in
sbi_hsm_hart_start_finish() cannot happen after 'start_ticket' has been
released.
Signed-off-by: Evgenii Shatokhin <e.shatokhin@yadro.com>
Reviewed-by: Anup Patel <anup@brainfault.org>
