To see the benefit of some optimization passes in LLVM for certain
architectures, I'd like to compile C or C++ code for a target where a C
compiler (SDCC) already exists.
I guess the easiest way would be to use clang -S -emit-llvm then the
optimization passes then the C backend (probably much easi
Is it possible to do a standalone build of clang, i.e. using LLVm from
the distro instead of building LLVM?
The instructions at http://clang.llvm.org/get_started.html only mention
building LLVM with clang.
At https://gist.github.com/kwk/9617978 there is a project (that I didn't
try yet) for a stan
On 18.09.2016 15:56, Philipp Klaus Krause via cfe-users wrote:
> Is it possible to do a standalone build of clang, i.e. using LLVm from
> the distro instead of building LLVM?
>
> The instructions at http://clang.llvm.org/get_started.html only mention
> building LLVM with cl
How can I make char unsigned by default in a target? I want to make my
target to behave by default as if -fno-signed-char had been specified.
TargetInfo has lots of members to specify properties of types, e.g.
WCharType (which can be overridden by options, such as
-fno-short-wchar). But I didn't s
In clang-3.8, I'm trying to have some data types aligned on 8-bit
boundaries, which works for everything except long long.
In my target, I've used:
IntAlign = 8;
LongAlign = 8;
LongLongAlign = 8;
FloatAlign = 8;
DataLayoutString =
"E-p:16:8:8-i1:8:8-i8:8:8-i16:8:8-i32:8:8-i64:8:8-f32:8:8-a:8:8";
Found another one that encountered this problem:
http://lists.llvm.org/pipermail/cfe-dev/2013-October/032554.html
Again no solution there.
Is anyone sucessfully using clang (preferably 3.8) with an alignment for
long long or double that is not 64 bits?
Philipp
_
On 22.10.2016 09:35, Philipp Klaus Krause via cfe-users wrote:
> Found another one that encountered this problem:
>
> http://lists.llvm.org/pipermail/cfe-dev/2013-October/032554.html
>
> Again no solution there.
>
> Is anyone sucessfully using clang (preferably 3.8) with an
On 06.11.2016 09:17, Hugh Perkins via cfe-users wrote:
> Hi,
>
> I have a project to rewrite llvm output as OpenCL,
> https://github.com/hughperkins/cuda-on-cl It currently contains lots of
> conditional branches, whch become /labels / gotos in the output OpenCL.
>
> What options are available t
