youben11
65e2e2f600
fix: correct reassociation in expand and collapse ops
2022-03-02 09:55:54 +01:00
Umut
d06e0c0a59
feat: enhance matmul operation to support 1-D and N-D inputs
2022-03-01 18:06:40 +03:00
Antoniu Pop
a4ae9b9541
fix(TestLib): fix error message typo.
2022-02-28 21:00:48 +00:00
Antoniu Pop
f3fca58efc
fix(runtime): add missing assert on context BSK.
2022-02-28 21:00:48 +00:00
Antoniu Pop
05059e8165
fix(ClientLib): correct the RuntimeContext in the argument list.
2022-02-28 21:00:48 +00:00
Antoniu Pop
f2bb077108
fix(TestLib): add missing test using a LUT/PBS.
2022-02-28 21:00:48 +00:00
Antoniu Pop
8ee456315e
fix(compiler): fix code generated for library header.
2022-02-28 21:00:48 +00:00
Antoniu Pop
ebb719cf60
fix(compiler): generate PIC code by default to allow generating shared libraries.
2022-02-28 21:00:48 +00:00
rudy
8b71e9d476
feat(Clientlib): separate client encryption and server computation
...
Resolve #200
2022-02-24 15:50:18 +01:00
youben11
4f19dce899
fix: adapt dialect tests after recent pipeline changes
...
new passes have affected the order of appearance of constants, so they
just needed to be put at the top in the expected IR
2022-02-24 09:44:26 +01:00
youben11
e82360a9fe
feat: support lowering of convolution end to end
2022-02-24 09:44:26 +01:00
youben11
86379096df
feat: lower FHELinalg.conv2d to linalg
...
This is currently lowering to our custom linalg conv operation, since
linalg ops doesn't support custom types. But as linalg will support
custom types in the future, we may want to lower to the native linalg op
instead
2022-02-24 09:44:26 +01:00
youben11
6d2f853c07
feat: support Conv2d in MANP
2022-02-24 09:44:26 +01:00
youben11
3668b2d73a
feat: add a Conv2d operation in FHELinalg
2022-02-24 09:44:26 +01:00
youben11
78596f899f
feat: add generated linalg conv operation
...
This has been generated using linalg tools, then put in their
appropriate locations. This is intended as a workaround since linalg
doesn't support tensors of custom types yet. Any conversion using this
added operation should be able to use the default operation from linalg
when it starts supporting tensor of custom types.
2022-02-24 09:44:26 +01:00
Quentin Bourgerie
7f0be306f0
test(compiler): Temporary disable test
...
This test is a bit artificial, let see if we actually want to fix it
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
61fefde35a
fix(compiler): Compilation gcc7
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
a068a728d3
chore(compiler): Fix missing dependencies in the build system
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
9627864d23
enhance(testlib): Fix the runtime testlib tools to handle the ciphertext bufferization and the new compiler concrete bufferized API
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
b1d6b7e653
test(compiler): TEMPORARY Disable tiling test since bufferization didn't yet support
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
0439ef47da
style(test): Formatting
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
de66044374
cleanup(test): Remove some useless references
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
4e8a9d1077
enhance(runtime): Fix the runtime tools to handle the ciphertext bufferization and the new compiler concrete bufferized API
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
8a9cce64e3
enhance(compiler): Add custom finalize bufferize pass to handle memref.tensor_load op
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
626493dda7
enhance(compiler): Lower from Concrete to BConcrete and BConcrete to C API call
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
b3368027d0
refactor(compiler): Move FHELinalg.zero to FHE.zero_tensor and add zero and sero_tensor in TFHE and Concrete dialects
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
c8b45cc853
style(compiler/client): fix the naming LweSize to LweDimension for key size
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
3330b2f855
fix(compiler): Take care of program with no input when creating client parameters
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
1f3944047f
feature(compiler): Introduce the BConcrete Dialect
2022-02-17 15:53:23 +01:00
Quentin Bourgerie
b950a0010d
chore(compiler/test): Move some test to end-to-end-jit-fhe
2022-02-17 15:53:23 +01:00
Umut
20a89b7b42
feat: implement concat operation
2022-02-15 16:52:07 +03:00
Umut
a1818a3fd9
feat: implement advanced sum operation
2022-02-08 17:49:58 +03:00
Antoniu Pop
dddad849c7
enhance(compiler): add --parallelize-loops and --parallelize-dataflow compile flags in addition to --parallelize which enables both.
2022-02-03 10:59:12 +00:00
Antoniu Pop
35e6966f95
enhance(dfr/runtime): throttle the number of HPX worker threads when using OpenMP for loop parallelism within tasks.
2022-02-03 10:59:12 +00:00
Antoniu Pop
ec633d57f6
fix(compiler): predicate OpenMP loop parallelization of Linalg on the --parallelize flag.
2022-02-03 10:59:12 +00:00
Quentin Bourgerie
f52965e22f
enhance(compiler): Lower linalg to parallel loops and instantiare parallel loops with openmp
2022-02-03 10:59:12 +00:00
Antoniu Pop
260768e9af
enhance(dfr): use linker wrapping of main in addition to start/stop; enable DFR+OpenMP.
2022-02-03 10:59:12 +00:00
Umut
4203e86998
feat: implement basic sum operation
2022-02-01 12:18:56 +03:00
Umut
a7c63a5494
fix: get rid of documentation mistakes and styling inconsistencies in FHELinalgOps.td
2022-02-01 12:18:56 +03:00
youben11
6d3af7fcda
chore: add installation targets in Makefile
...
MacOS require a way to find the runtime library when trying to build a
shared library at runtime, so we add a way to install it globally for
testing
2022-01-24 17:02:25 +00:00
youben11
71cccc6a89
fix(mac): add options for required libraries
2022-01-24 17:02:25 +00:00
youben11
faa0abea3e
fix: use appropriate cmd for dynamic lib generation
...
macos and linux require a different cmd
2022-01-24 17:02:25 +00:00
youben11
b8cfaeb3f4
fix: ignore narrowing errors on Mac
2022-01-24 17:02:25 +00:00
youben11
2926ea6a5e
fix: use appropriate type for optimizingTransformer func
...
the call does return an `std::function` and was being referenced using
an `llvm::function_ref`, which apparently with some optim on Mac was
referncing bad memory location
2022-01-24 17:02:25 +00:00
Umut
8198a79deb
feat: introduce maximum bit width constant, use it where appropriate, export it in python bindings
2022-01-24 17:26:37 +03:00
Andi Drebes
7fdeb61aa8
fix(compiler): Handle size_t explicitly in JITLambdaArgumentAdaptor::addArgument
...
On some systems, size_t does not alias any of the fixed-size, unsigned
integer types and therefore needs to be treated explicitly in
`JITLambdaArgumentAdaptor::addArgument` to prevent the function from
failing with an unknown argument type.
Closes issue #369 : Bug: MacOS tests failing on master due to
IntLambdaArgument<size_t>.
2022-01-10 16:57:37 +01:00
Andi Drebes
2d852165f6
fix(compiler): Make arrays passed to llvm::MutableArrayRef mutable
2022-01-07 12:03:13 +01:00
Andi Drebes
39b80bd5b2
fix(compiler): Remove unused variables in TFHEGlobalParametrization.cpp
2022-01-07 12:03:13 +01:00
Andi Drebes
7062052ca6
fix(compiler): Remove unused variables in lib/Dialect/TFHE/IR/TFHEOps.cpp
2022-01-07 12:03:13 +01:00
Andi Drebes
ce9d86014d
fix(compiler): Force same signedness of integers in _verifyGLWEIntegerOperator
2022-01-07 12:03:13 +01:00
