| Age | Commit message (Collapse) | Author |
|---|
|
This patch adds sscofpmf extension in pmu module which includes
following things.
1. Enable overflow irq when starting a counter.
2. Setting the correct event filters passed from supervisor.
3. Delegating the overflow interrupt to the supervisor.
4. Add RV32 support for sscofpmf.
Reviewed-by: Anup Patel <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
OpenSBI doesn't handle PMU counters for now.
Delegate the overflow counter to S-mode always.
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Sscofpmf ISA extension introduces PMU counter overflow and filtering support.
It introduces a read only `scountovf` csr that can be used to detect if
a hart supports this extension at runtime. However, this feature is only
useful if the hart already supports mcounteren and mcountinhibit.
Add a dynamic detection mechanism and boot time print message if sscofpmf
is present.
Reviewed-by: Anup Patel <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
The Sscofpmf extension introduces mhpmevent[h] csrs to handle filtering
/overflow bits in RV32. There is no way to read/write mcountinhibit
using mcountinhibit csr using a variable.
Updated the support to read/write mhpmevent[h] and mcountinhibit csr.
Reviewed-by: Anup Patel <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Moving the sbi_boot_print_general() call after the
sbi_platform_final_init() call allows to print devices initialized in the
latter.
To keep the overall print sequence the same also move
sbi_boot_print_domains().
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
The parameter passed to HFENCE.GVMA instruction in rs1 register
is guest physical address right shifted by 2 (i.e. divided by 4).
Unfortunately, we overlooked the semantics of rs1 registers for
HFENCE.GVMA instruction and never right shifted guest physical
address by 2. This issue did not manifest for hypervisors till
now because all H-extension implementations (such as QEMU, Spike,
Rocket Core FPGA, etc) we tried till now were conservatively
flushing everything upon any HFENCE.GVMA instruction.
This patch fixes GPA passed to __sbi_hfence_gvma_vmid_gpa()
and __sbi_hfence_gvma_gpa() functions.
Fixes: 331ff6a162c1 ("lib: Support stage1 and stage2 tlb flushing")
Reported-by: Ian Huang <[email protected]>
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Dong Du <[email protected]>
|
|
Although we have PRILX to help us print unsigned long without
considering the 32bit/64bit differences, there are still some
places using 08lx and 016lx manually --- leading to redundant code.
This commit fixes the issue by using PRILX all the time.
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
When we are doing opensbi development, we want to know the build time
and compiler info for debug purpose.
To enable this message, please add "BUILD_INFO=y", like:
```
make BUILD_INFO=y
```
NOTE: Using `BUILD_INFO=y` without specifying SOURCE_DATE_EPOCH will
violate "reproducible builds". So it's ONLY for development and debug
purpose, and should NOT be used in a product which follows "reproducible
builds".
Signed-off-by: Wei Fu <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
The SBI specification requires that sbi_system_reset() returns
SBI_ERR_INVALID_PARAM if reset_type or reset_reason are not valid.
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
Instead of saving context only for default non-retentive suspend,
we should save context for all non-retentive suspend types.
Fixes: 74756891cc35 ("lib: sbi: Implement SBI HSM suspend function")
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Dong Du <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
Let's make system_reset_check returning priority instead of only
true/false. In that case 0 - means not supported, and anything above
means priority that makes existing reset handlers being used in first
place, unless it is decided to lower their priority.
The handler with the most priority wins.
Signed-off-by: Nikita Shubin <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
To support different handlers for different types of resets, we are
adding a sbi_list of restart handlers.
Instead of sbi_system_reset_set_device we use
sbi_system_reset_add_device to reflect the actual meaning.
Signed-off-by: Nikita Shubin <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
We now have frequency of the timer device provided by the platform
support so we can emulate desired delay using a loop where the number
loop iterations are based on timer frequency.
This patch provides sbi_timer_delay_loop() for above purpose.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
We now have frequency in timer device instance provided by platform
so let's print timer frequency as part of the boot prints.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
We use BUG() macro in csr_read_num(), csr_write_num(), and
misa_string() functions for unhandled cases.
Signed-off-by: Xiang W <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
The original version of ctz will cause an endless loop, if the parameter
passed in is 0. This commit fixes this bug.
Signed-off-by: Xiang W <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
sbi_platform_get_features_str() uses sbi_snprintf() to construct the
features_str. However, it passes the wrong length value (i.e., the nfstr),
which should be (nfstr-offset) as the starting point of str (i.e.,
features_str + offset) changes.
This commit also checks the return value of snprintf, and handles the
corner case that the string buffer is full.
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
|
|
As other exported fifo functions, we should check whether the fifo is
valid in sbi_fifo_is_empty and sbi_fifo_is_full. To this end, this patch
changes the retval from bool to int, and the two functions will return
SBI_EINVAL in the case the fifo is invalid.
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
|
|
Avoid getting messages from multiple harts (using dprintf and printf)
concurrently with a spinlock serializaing calls to sbi_dprintf(),
sbi_printf() and sbi_puts()
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
Correcting the name of faults delegation CSR
%s/mfdeleg/medeleg
Signed-off-by: Rahul Pathak <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
commit 764a17d852a8 ("lib: sbi: Implement firmware counters") added
switch cases for CAUSE_LOAD_ACCESS and CAUSE_STORE_ACCESS. This caused
them to stop being redirected to U or S mode, as that is handled in the
default switch case. As a result, an error in userspace could cause the
system to hang. Fix this by allowing the acces fault case to fall
through to the default case.
Fixes: 764a17d852a8 ("lib: sbi: Implement firmware counters")
Signed-off-by: Samuel Holland <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
|
|
In sbi_ecall_rfence_handler, it will compare the funcid with
REMOTE_HFENCE_GVMA and REMOTE_HFENCE_VVMA_ASID. Later it check
whether the misa includes H-extension. This checking is incomplete
which misses REMOTE_HFENCE_GVMA_VMID and REMOTE_HFENCE_VVMA.
Fix the issue by updating the checking range.
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
|
|
No need to compare characters when the count turns to 0.
Fix the issue in sbi_strncmp.
Signed-off-by: Dong Du <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
The sbi_pmu_exit() crashes on systems not having MCOUNTINHIBIT csr
so to fix this we check SBI_HART_HAS_MCOUNTINHIBIT feature in
sbi_pmu_exit() and do nothing if it is not available.
Fixes: 13d40f21d588 ("lib: sbi: Add PMU support")
Signed-off-by: Anup Patel <[email protected]>
|
|
Follow the standard conventon for static function names:
All global functions should be start with sbi_<module name>_
All static functions should be start with <module name>_
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
RISC-V SBI v0.3 specification defines a set of firmware events that can
provide additional information about the current firmware context. All
of the firmware event monitoring are enabled now. The firmware
events must be defined as raw perf event with MSB set as specified in the
specification.
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
RISC-V SBI specfication 0.3 defines a PMU extension that allows supervisor
mode to start/stop/configure pmu related events. This patch implements
all of the functionality defined in the specification.
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
RISC-V SBI v0.3 specification defined a PMU extension to configure/start/stop
the hardware/firmware pmu events.
Implement PMU support in OpenSBI library. The implementation is agnostic of
event to counter mapping & mhpmevent value configuration. That means, it
expects platform hooks will be used to set up the mapping and provide
the mhpmevent value at runtime.
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Currently, csr_read/write_num functions are used to read/write PMP related
CSRs where CSR value is decided at runtime. Expand this function to include
PMU related CSRs as well.
Reviewed-by: Anup Patel <[email protected]>
Reviewed-by: Xiang W <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
This patch removes redundant print from sbi_boot_print_hart().
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Currently, all bits in mcountern are enabled unconditionally at boot time.
With SBI PMU extension, all the programmable counters should enabled only
during performance monitoring for a particular event. However, this is done
only if mcountinhibit is implemented because the supervisor mode can not
start/stop any event without mcountinhibit.
Similarly, supervisor should take care enabling scounteren which allows
U-mode to access programmable pmu counters. All the non-programmable ones
(CY, TM, IR) should be enabled in M-mode because some userspace may rely on
builtins such as __builtin_readcyclecounter. Supervisor OS can still disable
them during initial configuration.
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
RISC-V privilege specification allows the implementation to have less
than 64 bits.
Add a function to detect the number of implemented bits in mhpmcounter
dynamically at runtime.
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
RISC-V ISA specification v1.11 defined mcountinhibit CSR that allows
software to stop any counter from incrementing. The SBI PMU extension
depends on this CSR support in hardware.
Define mcountinhibit as a hart specific feature and detect it at runtime.
Reviewed-by: Anup Patel <[email protected]>
Signed-off-by: Atish Patra <[email protected]>
|
|
Don't assign an unused value to variable index.
Use operator '-=' where applicable.
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
Compiling wake_coldboot_harts() with GCC 11 and -Wextra yields:
lib/sbi/sbi_init.c:208:27:
error: comparison of integer expressions of different signedness:
‘int’ and ‘u32’ {aka ‘unsigned int’} [-Werror=sign-compare]
208 | for (int i = 0; i <= sbi_scratch_last_hartid(); i++) {
| ^~
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
Compiling sbi_tlb_entry_process() with GCC 11 and -Wextra yields:
lib/sbi/sbi_tlb.c: In function ‘sbi_tlb_process_count’:
lib/sbi/sbi_tlb.c:206:31:
error: comparison of integer expressions of different signedness:
‘u32’ {aka ‘unsigned int’} and ‘int’ [-Werror=sign-compare]
206 | if (deq_count > count)
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
In sbi_scratch_init() we determine the last hart. The index of the last
hart cannot exceed SBI_HARTMASK_MAX_BITS - 1. We should not initialize
last_hartid_having_scratch to a higher number to avoid buffer overflows
when using this value before calling sbi_scratch_init().
When allocating scratch memory in sbi_scratch_alloc_offset() we zero out
the allocated memory for all harts except for the last one. We should not
skip the last hart.
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
The parameter owner of function sbi_scratch_alloc_offset() is never used.
The scratch memory is small. We should not use it for debug information in
future. Hence eliminate the parameter.
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
We can merge region B onto region A only if base of region A is
aligned to region A order + 1.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
spin_lock_check already returned bool in the source file but not in the
header. With some toolchains that causes an error, as it should.
Because it and related functions all essentially return a bool, we can
use this opportunity to change them.
Signed-off-by: Daniel Schaefer <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
|
|
We extend the boot-time prints to show various devices provided
(or registered) by the platform support. This will help users
verify hardware devices available for OpenSBI firmwares.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
Instead of having hsm_start(), hsm_stop() and hsm_suspend()
callbacks in platform operations, it will be much simpler for
HSM driver to directly register these operations as a device
to the sbi_hsm implementation.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
Instead of having system_reset_check() and system_reset() callbacks
in platform operations, it will be much simpler for reset driver to
directly register these operations as a device to the sbi_system
implementation.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
Instead of having ipi_send() and ipi_clear() callbacks in
platform operations, it will be much simpler for ipi driver
to directly register these operations as a device to sbi_ipi
implementation.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
Instead of having timer_value(), timer_event_start(), and
timer_event_stop() callbacks in platform operations, it will
be much simpler for timer driver to directly register these
operations as device to the sbi_timer implementation.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
|
Instead of having console_putc() and console_getc() callbacks in
platform operations, it will be much simpler for console driver to
directly register these operations as device to the sbi_console
implementation.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
We now have sbi_domain_root_add_memregion() which allows platform
support to add root domain regions at boot-time so let's remove
the domains_root_regions() platform callback which was added
for this purpose.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
We should decrement root_memregs_count by one after merging two
memregions otherwise new memregion added after a merge will be
appended after last sentinel memregion of zero order.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Xiang W <[email protected]>
|
|
The sbi_domain_memregion_initfw() is no longer used outside
sbi_domain.c so let's make it a local function.
Signed-off-by: Anup Patel <[email protected]>
Reviewed-by: Xiang W <[email protected]>
Reviewed-by: Alistair Francis <[email protected]>
|
