current implementation incorrectly update CURRENT_REGS to interrupt context if
trigger nested interrupt, (e.g, hard fault occurs during interrupt handling)
this would ambiguous for programs using CURRENT_REGS, this patch will prohibit
the update of CURRENT_REGS on nested interrupt handling
Signed-off-by: chao.an <anchao@xiaomi.com>
and removed two tcp_send_txnotify() calls from tcp_sendfile (they are not needed anymore).
As a result, the TX throughput of both the tcp_send_buffered and tcp_send_unbuffered
is significantly boosted in case of TUN/TAP network device.
Basically, mirror the following two commits from modlib.
It's shame we have two copies of elf loaders.
```
commit 51490bad55
Author: YAMAMOTO Takashi <yamamoto@midokura.com>
Date: Wed Apr 14 17:07:39 2021 +0900
modlib: Implement sh_addralign handling
I've seen a module with 16 bytes .rodata alignment for xmm operations.
It was getting SEGV on sim/Linux because of the alignment issue.
The same module binary seems working fine after applying this patch.
Also, tested on sim/macOS and esp32 on qemu,
using a module with an artificially large alignment. (64 bytes)
```
```
commit 418e11b8b3
Author: YAMAMOTO Takashi <yamamoto@midokura.com>
Date: Thu Apr 15 11:33:48 2021 +0900
modlib: Always use separate allocation for text and data
Pros:
* Reduce code differences
* Smaller allocations for !CONFIG_ARCH_USE_MODULE_TEXT
Cons:
* Likely to use more memory for !CONFIG_ARCH_USE_MODULE_TEXT in total
Tested with:
* sim:module on macOS
* esp32-devkit:nsh + CONFIG_MODULE on qemu
* lm3s6965-ek:qemu-protected + CONFIG_EXAMPLES_SOTEST on qemu
```
Some PPG devices have 4 ADCs to output quad-channel PPG values while some of them only have 2 ADCs to output dual-channel PPG.
To deal the case above, the type PPGD(PPG of Dual-channel) takes the place of former type PPG, which also have 2-channel PPG outputs. Type PPGQ (PPG of Quad-channel) is new for 4-ADC-PPG. Both types have contained new data "gain" to indicate ADC gains of each PPG channel, for the reason that the gains may vary during automatical optical adjustments.
Signed-off-by: liucheng5 <liucheng5@xiaomi.com>
The image must be placed into:
boards/risc-v/mpfs/icicle/include/boot_romfsimg.h
The image is mounted by mpfs_bringup, which is run by the application
itself, or by board_late_initialize() in the case when
CONFIG_BOARD_LATE_INITIALIZE is defined, e.g. with CONFIG_BUILD_KERNEL
According to RFC 5681 (3.2) the TCP Fast Retransmit algorithm should start
if the threshold of 3 duplicate ACKs is reached.
Thus the threshold should be a constant, not an integer option.
Perf timer interface generally uses the hardware cycle counter
provided by the arch chip directly(such as DWT_CYCCNT(cortex-m)),
CYCCNT is a free running counter and counting upwards.
It wraps around to 0 on overflow.
Signed-off-by: chao.an <anchao@xiaomi.com>
- Add test for mode support, which is architecture dependent
- Add tests for address alignment and region size
- Add option to query for access rights
- The function goes through every PMP entry and tests if an address
range from [base, base+size] has been configured for desired
access rights.
- If several PMP entries match the range and access rights, the
information is combined
- End result is either no access, a partial match was found, or a full
match was found. Details about the partial match are not provided.
The intent for testing access rights and not just blindly applying them
is a case where they are already set in e.g. a bootloader. In this
case, nothing should be done, unless the configuration does not match,
in which case the software must not continue further.
NOTE: THIS ONLY WORKS WHEN KERNEL RUNS IN M-MODE FOR NOW
This frees the PMP for other use, e.g. HART memory separation.
The page tables are statically allocated, 1 per level.
This feature is now behind CONFIG_MPFS_USE_MMU_AS_MPU, because
only the MPFS target supports this (others are not tested).
If the MMU is used for memory separation within a HART, the PMP must
still be configured to allow user access to the memory mapped for the
HART, because PMP *rekoves* access by default. At this point all of
the user memory as well as the kernel RAM are opened.
A more flexible solution for PMP configuration will follow.
The old implementation used the default ld.script for the kernel side
which did not obey the memory.ld limits whatsoever.
Also, provide the user space addresses from the linker script to get rid
of the pre-processor macros that define (incorrect) default values for
the user space composition.
Summary:
- I noticed that make export does not work with swama5d4-ek:knsh
- This commit fixes this issue.
- NOTE: apps/Makefile also needs to be updated.
Impact:
- CONFIG_BUILD_KERNEL=y only
Testing:
- Build (make and make export) with sama5d4-ek:knsh
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
- remove some unneeded includes
- tab/space fixes
- change DEBUGPANIC to PANIC in noreturn function, otherwise it compiles only in DEBUG builds
Signed-off-by: Jukka Laitinen <jukkax@ssrc.tii.ae>
The following options need to be taken in account while determining
the proper hart_index2id -table:
CONFIG_MPFS_HART1_SBI
CONFIG_MPFS_HART1_SBI
CONFIG_MPFS_HART3_SBI
CONFIG_MPFS_HART4_SBI
Unused harts should be marked with -1. Hart0 is never used so it
stays at -1.
Signed-off-by: Eero Nurkkala <eero.nurkkala@offcode.fi>
