mirror of
https://github.com/AsahiLinux/u-boot
synced 2025-02-26 04:17:09 +00:00
On 32-core platform, hart31 gets stuck at secondary_hart_loop as the corresponding enable bit is not set in enable_ipi(). We should program the next word (0x2f84) which is assigned as the enable register of hart31. It should be done in the same way when we invoke riscv_send_ipi() to trigger software interrupt on hart31. The following diagram shows the enable bits of the fixed PLICSW scheme. Pending regs: 0x1000 x---0---0---0---0------0---0 Pending hart ID: 0 1 2 3 ... 30 31 Interrupt ID: 0 1 2 3 4 ... 31 32 | | | | | | | Enable regs: 0x2000 x---1---0---0---0-...--0---0---> hart0 | | | | | | | 0x2080 x---0---1---0---0-...--0---0---> hart1 | | | | | | | 0x2100 x---0---0---1---0-...--0---0---> hart2 | | | | | | | 0x2180 x---0---0---0---1-...--0---0---> hart3 . . . . . . . . . . . . . . . . . . . . . 0x2f00 x---0---0---0---0-...--1---0---> hart30 | | | | | | | 0x2f80 x---0---0---0---0-...--0---1---> hart31 <-------- word 0 -------><--- word 1 ---> This patch includes some cleanups to macros/functions. Fixes: ebf11273220a ("riscv: andes: Rearrange Andes PLICSW to single-bit-per-hart strategy") Signed-off-by: Yu Chien Peter Lin <peterlin@andestech.com> Reviewed-by: Randolph <randolph@andestech.com> |
||
---|---|---|
.. | ||
aclint_ipi.c | ||
andes_plicsw.c | ||
asm-offsets.c | ||
boot.c | ||
bootm.c | ||
cache.c | ||
crt0_riscv_efi.S | ||
elf_riscv32_efi.lds | ||
elf_riscv64_efi.lds | ||
fdt_fixup.c | ||
image.c | ||
interrupts.c | ||
Makefile | ||
memcpy.S | ||
memmove.S | ||
memset.S | ||
reloc_riscv_efi.c | ||
reset.c | ||
sbi.c | ||
sbi_ipi.c | ||
semihosting.S | ||
setjmp.S | ||
sifive_cache.c | ||
smp.c | ||
spl.c | ||
strcmp_zbb.S | ||
strlen_zbb.S | ||
strncmp_zbb.S |