Commit Graph

8195 Commits

Author SHA1 Message Date
Sebastien Bourdeauducq 17b9d2fc5a flake: add KC705 HITL test 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 5e2664ae7e flake: add openocd 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 64ce7e498b flake: make board package a Python package 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 952acce65b flake: build board package on Hydra 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 7ae4b2d9bb flake: update dependencies 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq ce0964e25f flake: fix cargo sha256 2021-09-10 13:25:12 +08:00
occheung 4fab267593 cargo: std dependency hack 2021-09-10 13:25:12 +08:00
occheung dcbd9f905c cargo: use cargo xbuild 2021-09-10 13:25:12 +08:00
occheung 9f6b3f6014 firmware: clarify target triple
The lack of compressed instruction support can be inferred from the target triple, literally.
2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 9697ec33eb flake: update dependencies 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq eee80c7697 flake: use improved Rust support in nixpkgs 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq b7efb2f633 flake: remove outdated comment 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 9ee03bd438 flake: reenable lit test 2021-09-10 13:25:12 +08:00
occheung 4619a33db4 test: remove broken array return tests
Removed test cases that do not respect lifetime/scope constraint.
See discussion in artiq-zynq repo: M-Labs/artiq-zynq#119
Referred to the patch from @dnadlinger. 5faa30a837
2021-09-10 13:25:12 +08:00
occheung 5985f7efb5 syscall: lower nowrite to inaccessiblememonly
In the origin implementation, the `nowrite` flag literally means not writing memory at all.
Due to the usage of flags on certain functions, it results in the same issues found in artiq-zynq after optimization passes. (M-Labs/artiq-zynq#119)
A fix wrote by @dnadlinger can resolve this issue. (c1e46cc7c8)
2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 6db7280b09 flake: board package WIP 2021-09-10 13:25:12 +08:00
occheung d8ac429059 dyld: streamline lib.rs
Only riscv32 is supported anyway, no need to have excessive architecture check.
2021-09-10 13:25:12 +08:00
occheung 798774192d slave_fpga/bootloader: read in little endian 2021-09-10 13:25:12 +08:00
occheung eecd825d23 firmware: suppress warning 2021-09-10 13:25:12 +08:00
occheung 1da0554a49 pcr: purge 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 035d15af9d flake: clean up vivado, add installer environment 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 9addd08587 flake: fetch MiSoC submodules 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 3e09e48152 flake: set up Vivado 2021-09-10 13:25:12 +08:00
occheung 5d0a8cf9ac llvm_ir_gen: fix indent 2021-09-10 13:25:12 +08:00
occheung 70507e1b72 Cargo.lock: update 2021-09-10 13:25:12 +08:00
occheung c113cd6bf5 libfringe: bump 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 251cd4dcc6 flake: update dependencies 2021-09-10 13:25:12 +08:00
occheung 61b0170a12 firmware: purge or1k 2021-09-10 13:25:12 +08:00
occheung af263ffe1f ksupport: fix rpc, cache signature (FFI)
The reason of the borrow stuff is explained in M-Labs/artiq-zynq#76 (artiq-zyna repo).
As for `cache_get()`, compiler will perform stack allocation to pre-allocate the returned structure, and pass to cache_get alongside the `key`.
However, ksupport fails to recognize the passed memory, so it will always assume the passed memory as the key.
2021-09-10 13:25:12 +08:00
occheung a833974b50 analyzer: fix endianness 2021-09-10 13:25:12 +08:00
occheung d623acc29d llvm_ir_gen: fix now with now_pinning & little-endian target 2021-09-10 13:25:12 +08:00
occheung 8fa47b8119 rpc: enforce alignment 2021-09-10 13:25:12 +08:00
occheung de0f2d4a28 firmware: adopt endianness protocol in artiq-zynq
Related:
artiq-zynq: M-Labs/artiq-zynq#126
artiq: #1588
2021-09-10 13:25:12 +08:00
occheung 9afe63c08a ksupport: fix proto_artiq dependency 2021-09-10 13:25:12 +08:00
occheung 29a2f106d1 ksupport: replace asm with llvm_asm 2021-09-10 13:25:12 +08:00
occheung b30ed75e69 kernel.ld: load elf header and prog headers
ld.lld has a habit of not putting the headers under any load sections.
However, the headers are needed by libunwind to handle exception raised by the kernel.
Creating PT_LOAD section with FILEHDR and PHDRS solves this issue. Other PHDRS are also specified as linkers (not limited to ld.lld) will not create additional unspecified headers even when necessary.
2021-09-10 13:25:12 +08:00
occheung 279593f984 ksupport.ld: merge sbss with bss 2021-09-10 13:25:12 +08:00
occheung 1ba8c8dfee runtime: remove irq again 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 942bd1a95d flake: add hydraJobs 2021-09-10 13:25:12 +08:00
occheung 3d629006df makefiles: revert byte-swaps 2021-09-10 13:25:12 +08:00
occheung 7542105f0f board_misoc: remove pcr
VexRiscv seems to not support additional hardware performance counter, at least I have not seen any documentation on how to use it.
2021-09-10 13:25:12 +08:00
occheung 01ca114c66 runtime: remove irq dependency 2021-09-10 13:25:12 +08:00
occheung 36171f2c61 runtime: remove inaccurate sp on panic 2021-09-10 13:25:12 +08:00
occheung 01e357e5d3 ksupport.ld: reduce load section alignment 2021-09-10 13:25:12 +08:00
occheung f77b607b56 compiler: generate symbols 2021-09-10 13:25:12 +08:00
occheung 1293e0750e ld, makefiles: use ld.lld 2021-09-10 13:25:12 +08:00
occheung fc42d053d9 kernel: use vexriscv 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 9adab6c817 flake: add devshell 2021-09-10 13:25:12 +08:00
Sebastien Bourdeauducq 8c468d0346 flake: switch to nightly rust with mozilla overlay 2021-09-10 13:25:12 +08:00
occheung 1b516b16e2 targets: default to vexriscv cpu 2021-09-10 13:25:12 +08:00