Commit Graph

611 Commits

Author SHA1 Message Date
Masayuki Ishikawa
1a9e7efde5 smp: Remove CONFIG_SMP_IDLETHREAD_STACKSIZE
Summary:
- The CONFIG_SMP_IDLETHREAD_STACKSIZE was introduced to optimize
  the idle stack size for other than CPU0
- However, there are no big differences between the idle stacks.
- This commit removes the config to simplify the kernel code

Impact:
- All SMP configurations

Testing:
- Tested with ostest with the following configs
- spresense:smp, spresense:rndis_smp
- esp32-devkitc:smp (QEMU), maix-bit:smp (QEMU)
- sabre-6quad:smp (QEMU), sabre-6quad:netnsh_smp (QEMU)
- raspberrypi-pico:smp, sim:smp (x86_64)

Signed-off-by: Masayuki Ishikawa <asayuki.Ishikawa@jp.sony.com>
2021-04-19 21:46:39 -05:00
Abdelatif Guettouche
c1b0ee436c arch/xtensa/src/esp32/Kconfig: Make bank switching default to disabled.
This config is only useful when there is a > 4MB PSRAM and thus needs to
be selected by the user explicitly.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-04-19 07:48:35 -05:00
Xiang Xiao
2335b69120 arch: Allocate the space from the beginning in up_stack_frame
arch: Allocate the space from the beginning in up_stack_frame

and modify the affected portion:
1.Correct the stack dump and check
2.Allocate tls_info_s by up_stack_frame too
3.Move the stack fork allocation from arch to sched

Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2021-04-16 12:41:41 +09:00
Xiang Xiao
8640d82ce0 arch: Rename g_intstackbase to g_intstacktop
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2021-04-16 12:41:41 +09:00
YAMAMOTO Takashi
3806803a7a arch/xtensa/src/esp32/esp32_user.c: Implement L16SI emulation
I don't know why this was not necessary before.
Probably I was just lucky about the combination of configs.
Or maybe some of recent changes happened to make the compiler
to use the instruction.

```
400d38f0 <mm_givesemaphore>:
400d38f0:       004136          entry   a1, 32
400d38f3:       228c            beqz.n  a2, 400d38f9 <mm_givesemaphore+0x9>
400d38f5:       0228            l32i.n  a2, a2, 0
400d38f7:       52cc            bnez.n  a2, 400d3900 <mm_givesemaphore+0x10>
400d38f9:       fea0b2          movi    a11, 254
400d38fc:       000306          j       400d390c <mm_givesemaphore+0x1c>
400d38ff:       00              .byte 00
400d3900:       019232          l16si   a3, a2, 2
400d3903:       feebe5          call8   400d27c0 <getpid>
400d3906:       0813a7          beq     a3, a10, 400d3912 <mm_givesemaphore+0x22>
400d3909:       05a1b2          movi    a11, 0x105
400d390c:       f241a1          l32r    a10, 400d0210 <_stext+0x1f0>
400d390f:       ff23e5          call8   400d2b4c <_assert>
400d3912:       1288            l32i.n  a8, a2, 4
400d3914:       0828a6          blti    a8, 2, 400d3920 <mm_givesemaphore+0x30>
400d3917:       880b            addi.n  a8, a8, -1
400d3919:       1289            s32i.n  a8, a2, 4
400d391b:       000606          j       400d3937 <mm_givesemaphore+0x47>
400d391e:       00              .byte 00
400d391f:       00              .byte 00
400d3920:       ffaf82          movi    a8, -1
400d3923:       015282          s16i    a8, a2, 2
400d3926:       00a082          movi    a8, 0
400d3929:       016282          s32i    a8, a2, 4
400d392c:       02ad            mov.n   a10, a2
400d392e:       feb125          call8   400d2440 <sem_post>
400d3931:       19a1b2          movi    a11, 0x119
400d3934:       fd4a96          bltz    a10, 400d390c <mm_givesemaphore+0x1c>
400d3937:       f01d            retw.n
400d3939:       000000          ill
```
2021-04-15 12:18:52 +01:00
YAMAMOTO Takashi
a28de1d681 arch/xtensa/src/esp32/esp32_user.c: Fix S16I/L16LU emulation
I misunderstood how imm8 is used to calculate the address.
2021-04-15 12:18:52 +01:00
YAMAMOTO Takashi
51490bad55 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)
2021-04-14 21:17:07 -05:00
Alan Carvalho
ac5fb7d701 esp32: Fix GPIO Pull-Up/Pull-Down using RTC GPIO
Some ESP32 GPIO pins (2, 4, 12, 13, 25, 27, 32) weren't accepting
pull-up/pull-down resistors. These pins are RTC GPIO pins and need
to have pull-up/pull-down configured in the RTC registers.

Co-authored-by: Gustavo Henrique Nihei <gustavo.nihei@espressif.com>
2021-04-11 14:36:02 +01:00
Xiang Xiao
3f67c67aaf arch: Fix the stack boundary calculation and check
All supported arch uses a push-down stack:
The stack grows toward lower addresses in memory. The stack pointer
register points to the lowest, valid working address (the "top" of
the stack). Items on the stack are referenced as positive(include zero)
word offsets from sp.
Which means that for stack in the [begin, begin + size):
1.The initial SP point to begin + size
2.push equals sub and then store
3.pop equals load and then add

Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2021-04-10 08:39:54 -07:00
Xiang Xiao
0fdde5be26 arch/esp32: Fix error: Mixed case identifier found
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
2021-04-10 12:00:06 +01:00
Gustavo Henrique Nihei
4d4250fcca xtensa/esp32: Improve SPI polling to use the entire HW buffer 2021-04-08 23:36:28 -05:00
Matias N
ab206687bb Replace wrong inclusion of sys/errno.h (toolchain provided) with errno.h 2021-04-07 21:27:06 -05:00
Alin Jerpelea
cb2ecefbf1 arch: xtensa: fix nxstyle errors
Fix for errors reported by nxstyle tool

Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
2021-04-07 21:21:51 -05:00
Matias N
d88b5aac97 esp32: move common XTAL and RUN_IRAM configs to ESP32 KConfig 2021-04-07 21:45:48 +01:00
Alan Carvalho de Assis
bac84de45f esp32c3: Add support to RNG driver
Co-authored-by: Gustavo Henrique Nihei <gustavo.nihei@espressif.com>
Co-authored-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-04-03 07:20:03 -05:00
Alan C. Assis
18f88c35fc esp32: Fix small typo that will trigger an error when IPv6 is enabled 2021-04-03 00:53:02 -05:00
Alin Jerpelea
778f050102 arch: xtensa: Author Gregory Nutt: update licenses to Apache
Gregory Nutt has submitted the SGA and we can migrate the licenses
 to Apache.

Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
2021-04-02 03:14:31 -05:00
Gustavo Henrique Nihei
2d0e690803 xtensa/esp32: Refactor register access functions on SPI driver 2021-04-01 17:13:55 -03:00
Alin Jerpelea
3d96d5f2ce arch: esp32: Mixed Case identifier fix
Fix Mixed Case Identifier reported by nxstyle

Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
2021-04-01 12:13:12 -05:00
Alin Jerpelea
4e26e39ffe arch: xtensa: Espressif Systems: update licenses to Apache
Espressif Systems has submitted the SGA and we can migrate the licenses
 to Apache.

Gregory Nutt has submitted the SGA and we can migrate the licenses
 to Apache.

Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
2021-04-01 12:13:12 -05:00
Matias N
8ed2bb8eb5 esp32: remove unneeded "INFO" and "ERROR" prefixes to syslog calls 2021-03-31 07:37:18 -03:00
Matias N
a5a435e98f esp32: remove extra initial newline on syslog call 2021-03-31 10:04:42 +01:00
Gustavo Henrique Nihei
77c5995f93 xtensa/esp32: Use essential boolean expressions on condition statements 2021-03-30 01:23:02 -05:00
Gustavo Henrique Nihei
5e8eb420b7 xtensa/esp32: Fix MISO/MOSI data length field configuration
Both fields were being configured with the total number of remaining
bytes instead of the number of bytes actually bound to DMA descriptors.
2021-03-30 01:23:02 -05:00
Gustavo Henrique Nihei
b4dbae1b10 xtensa/esp32: Commit setbits configuration before SPI transaction
The motivation is to avoid consistency issues when using Mixed Mode
(i.e. Polling and Interrupt/DMA transfers being used interchangeably)
2021-03-30 01:23:02 -05:00
Gustavo Henrique Nihei
4d877abf3f xtensa/esp32: Avoid incrementing a NULL pointer for RX buffer 2021-03-30 01:23:02 -05:00
Gustavo Henrique Nihei
a27d5b1063 xtensa/esp32: Remove useless pointer check in SPI DMA exchange 2021-03-30 01:23:02 -05:00
Brennan Ashton
0a3b20e546 syslog: Drop extra carriage return from syslog calls 2021-03-28 21:24:00 -05:00
Gustavo Henrique Nihei
96037f01d5 xtensa/esp32: Clean up unused include headers from DMA driver 2021-03-26 23:40:44 -05:00
Gustavo Henrique Nihei
d3342795a8 xtensa/esp32: Fix wrong math round operation on DMA init 2021-03-26 23:40:44 -05:00
Gustavo Henrique Nihei
eb505ed866 xtensa/esp32: Fix DMA burst mode being unintendedly disabled 2021-03-26 23:39:53 -05:00
Sara Souza
59313c86d1 xtensa/esp32: Adds oneshot timer driver. 2021-03-24 16:01:26 -03:00
chenwen
f54aef9977 xtensa/esp32: Support esp32 wireless ioctl cmd 2021-03-23 16:29:52 -03:00
Abdelatif Guettouche
fcafacb9a3 esp32_allocateheap.c: Adjust the region of the heap coming from the
external memory when a BSS section is allowed to reside there.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-23 16:39:01 +09:00
Abdelatif Guettouche
cc23bdeca4 boards/xtensa/esp32: Add a section in external memory to hold some BSS
data.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-23 16:39:01 +09:00
YAMAMOTO Takashi
37300a43a5 esp32_part_ioctl: Return -ENOTTY for unknown commands
It's traditional to use ENOTTY for this purpose.
Littlefs seems to rely on this behavior for BIOC_FLUSH.

Also, drop the log level.
2021-03-22 19:49:27 -07:00
Gustavo Henrique Nihei
e4efa9dfa7 xtensa/esp32: Fix interrupt flag configuration for DMA transfers
Previously SPI interrupts were enabled on DMA initialization. But since
the addition of SPI Mixed mode it created a side-effect, breaking
polling transfers. So now interrupts are enabled before the DMA
transactions and disabled once they are finished.
Furthermore, the transaction done flag is also cleared before a new
transaction starts.
2021-03-21 00:16:59 -07:00
Gustavo Henrique Nihei
20d24fe148 xtensa/esp32: Fix esp32_spi_setbits for Polling when DMA is also enabled
Commit 6382b2ba introduced the possibility of using SPI in Mixed mode,
i.e. performing SPI transfers via both polling and interrupts. However,
setbits was only applying the configuration if DMA was not enabled.
2021-03-21 00:16:59 -07:00
Gustavo Henrique Nihei
27e2da33b4 xtensa/esp32: Fix buffer size word-alignment for DMA transfers 2021-03-20 19:23:44 -07:00
Gustavo Henrique Nihei
bfc551484a xtensa/esp32: Clean up esp32_dma_init code
Removed "isrx" parameter whose only purpose is to trigger an assertion
on DEBUG builds. Also performed a minor refactor.
2021-03-20 19:23:44 -07:00
Gustavo Henrique Nihei
dc7a0b0a5c xtensa/esp32: Use Polling instead of DMA for transfers below threshold
Also refactored code to remove a confusing duplicate "dma_chan" field
which had the same purpose of the "use_dma" boolean.
2021-03-19 23:13:32 -07:00
Abdelatif Guettouche
27d5c9340a esp32_allocateheap.c: Don't allocate the ROM CPU regions the same way in
QEMU, the image is different.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-18 11:28:36 +09:00
Jiuzhu Dong
e96c8b9283 fs: allocate file/socket dynamically
Change-Id: I8aea63eaf0275f47f21fc8d5482b51ffecd5c906
Signed-off-by: Jiuzhu Dong <dongjiuzhu1@xiaomi.com>
2021-03-17 06:46:42 -07:00
Dong Heng
b2f5031e96 xtensa/esp32: Refactor ESP32 WiFi driver to support station and softAP coexistence 2021-03-16 10:20:59 -03:00
Abdelatif Guettouche
28160823b6 arch/xtensa/esp32: ~6KB of memory at address 0x3ffae6f0 is not used by
the ROM bootloader, add that to the heap as well.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00
Abdelatif Guettouche
8389e83742 esp32/memory_layout.h: Update the layout taking under consideration the
changes to the heap regions and to the internal heap.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00
Abdelatif Guettouche
9cfc30fa85 memory_layout.h: Fix the start of region2 when a QEMU image generation
is enabled.

That region is technically part of the PRO CPU and we should be able to
allocate it early.  However, QEMU uses a slightly different bootloader
image that uses the same part for both CPU.  So, when APP CPU starts
during the SMP bring up it will corrupt some data.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00
Abdelatif Guettouche
7fbc350589 xtensa/esp32: Warn about unused memory regions.
In case CONFIG_MM_REGIONS doesn't include all the available memory
regions the user will have a warning to increase it.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00
Abdelatif Guettouche
5c7d041b91 arch/xtensa/esp32: In SMP case move the internal memory to region 3.
Region 2 is only 15KB in SMP, so we don't have enough memory to play
with.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00
Abdelatif Guettouche
cba44928d2 arch/xtensa/esp32: Part of the ROM regions in middle of DRAM are not
used, retrieve them as heap.

Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
2021-03-16 16:22:08 +09:00