forked from M-Labs/nac3
1
0
Fork 0
Commit Graph

33 Commits

Author SHA1 Message Date
David Mak c93ad152d7 core: Codegen for ellipsis expression as NotImplemented
A lot of refactoring was performed, specifically with relaxing
expression codegen to return Option in case where ellipsis are used
within a subexpression.
2023-12-08 18:00:51 +08:00
David Mak c2ab6b58ff artiq: Implement `with legacy_parallel` block 2023-11-04 13:42:44 +08:00
David Mak 0a84f7ac31 Add CodeGenerator::gen_block and refactor to use it 2023-11-04 13:42:44 +08:00
David Mak 9518d3fe14 artiq: Fix timeline not resetting upon exiting sequential block 2023-10-30 14:04:53 +08:00
David Mak cbd333ab10 artiq: Extract `parallel` block timeline utilities 2023-10-30 14:04:53 +08:00
David Mak 65d6104d00 artiq: Improve IR value naming and add documentation 2023-10-30 14:04:53 +08:00
David Mak 8373a6cb0f artiq: Use gen_block when generating "with sequential" 2023-10-30 14:04:53 +08:00
David Mak c7de22287e core: Fix restoration of stack address
All allocas for temporary objects are now placed in the beginning of the
function. Allocas for on-temporary objects are not modified because
these variables may appear in a loop and thus must be uniquely
represented.
2023-10-06 11:34:23 +08:00
David Mak 31dcd2dde9 core: Use i8 for boolean variable allocation
In LLVM, i1 represents a 1-byte integer with a single valid bit; The
rest of the 7 upper bits are undefined. This causes problems when
using these variables in memory operations (e.g. memcpy/memmove as
needed by List slicing and assignment).

We fix this by treating all local boolean variables as i8 so that they
are well-defined for memory operations. Function ABIs will continue to
use i1, as memory operations cannot be directly performed on function
arguments or return types, instead they are always converted back into
local boolean variables (which are i8s anyways).

Fixes #315.
2023-09-25 15:42:07 +08:00
David Mak 6dfc43c8b0 core: Add name to build_gep_and_load 2023-09-22 17:16:29 +08:00
David Mak f1b0e05b3d core: Rename IR variables
Because it is unclear which variables are expressions and
subexpressions, all variables which are previously anonymous are named
using (1) the control flow statement if available, (2) the possible name
of the variable as inferred from the variable name in Rust, and (3) the
"addr" prefix to indicate that the values are pointers. These three
strings are joint together using '.', forming "for.i.addr" for instance.
2023-09-06 14:02:15 +08:00
David Mak ff23968544 core: Add name parameter to gen_{var_alloc,store_target}
This allows variables in the IR to be assigned a custom name as opposed
to names with a default prefix.
2023-09-06 11:00:02 +08:00
David Mak d37287a33d Cargo: Update dependencies 2023-09-04 10:43:57 +08:00
Sebastien Bourdeauducq c1c45373a6 update dependencies 2023-01-12 19:31:03 +08:00
ychenfo 6c485bc9dc nac3artiq: skip attribute writeback for option
option types do not have any fields to be written back to the host so it is ok to skip. If we do not skip, there will be error when getting the value of it since it can be `none`, whose type is not concrete
2022-04-10 01:28:30 +08:00
ychenfo 089bba96a3 nac3artiq: get_obj_value take an additional argument for expected type 2022-04-10 01:28:30 +08:00
ychenfo 0e0871bc38 nac3core, artiq: to_basic_value_enum takes an argument indicating the expected type 2022-04-10 01:28:22 +08:00
ychenfo 6ab73a223c nac3core/artiq: support default param of option type 2022-03-30 04:05:47 +08:00
pca006132 bf067e2481 nac3artiq: implement attribute writeback
We will only writeback attributes that are supported by the current RPC
implementation: primitives, tuple and lists of lists... of primitives.
2022-03-26 20:13:58 +08:00
pca006132 94aac16cc5 nac3artiq: fixed RPC codegen for lists 2022-03-10 16:48:28 +08:00
pca006132 323d77a455 nac3artiq: improve error message for out of range error 2022-02-28 23:09:14 +08:00
pca006132 f97f93d92c applied rustfmt and clippy auto fix 2022-02-21 18:27:46 +08:00
pca006132 d9cb506f6a nac3core: refactored for better error messages 2022-02-21 18:24:19 +08:00
pca006132 bf52e294ee nac3artiq: RPC support 2022-02-12 22:50:32 +08:00
pca006132 fa04768a77 redo "nac3core: fix #84"
This reverts commit 86005da8e1.
2021-12-27 22:56:26 +08:00
Sebastien Bourdeauducq 86005da8e1 Revert "nac3core: fix #84"
This reverts commit 0902d8adf4.
2021-12-26 08:35:27 +08:00
Sebastien Bourdeauducq 9e3f75255e update inkwell. Closes #67 2021-12-25 22:17:06 +08:00
pca006132 0902d8adf4 nac3core: fix #84 2021-12-23 15:26:48 +08:00
pca006132 ba08deada6 nac3core: refactor codegen 2021-11-20 19:50:25 +08:00
Sebastien Bourdeauducq bf7e2c295a integrate nac3parser 2021-11-03 17:11:00 +08:00
Sebastien Bourdeauducq bc0f82cad8 Revert "nac3artiq/codegen: fixed smax problem"
We have LLVM 12 now and can use the intrinsic.

This reverts commit 98d9f73afb.
2021-11-02 14:00:28 +08:00
pca006132 98d9f73afb nac3artiq/codegen: fixed smax problem
It turns out the smax intrinsic I use is a new one that is not supported
in LLVM11. Now implemented with signed integer compare and select.
2021-11-02 11:10:21 +08:00
pca006132 84c5201243 with parallel/sequential support
Behavior of parallel and sequential:
Each function call (indirectly, can be inside a sequential block) within a parallel
block will update the end variable to the maximum now_mu in the block.
Each function call directly inside a parallel block will reset the timeline after
execution. A parallel block within a sequential block (or not within any block) will
set the timeline to the max now_mu within the block (and the outer max now_mu will also
be updated).

Implementation: We track the start and end separately.
- If there is a start variable, it indicates that we are directly inside a
parallel block and we have to reset the timeline after every function call.
- If there is a end variable, it indicates that we are (indirectly) inside a
parallel block, and we should update the max end value.

Note: requires testing, it is difficult to inspect the output IR
2021-10-31 17:16:21 +08:00