diff options
author | Dietmar Maurer <dietmar@mono-cvs.ximian.com> | 2002-03-15 11:23:34 +0300 |
---|---|---|
committer | Dietmar Maurer <dietmar@mono-cvs.ximian.com> | 2002-03-15 11:23:34 +0300 |
commit | 04bbd9582f186151f25730a31cda0341df276148 (patch) | |
tree | 38ed7d3d0b270759f07e6abeef458dc65c50fe46 /docs/jit-thoughts | |
parent | f1a186b65637f5dbd216e0d28f30f78ad2cb8376 (diff) |
documentation updates
svn path=/trunk/mono/; revision=3126
Diffstat (limited to 'docs/jit-thoughts')
-rw-r--r-- | docs/jit-thoughts | 65 |
1 files changed, 4 insertions, 61 deletions
diff --git a/docs/jit-thoughts b/docs/jit-thoughts index 90cf4785523..cb50d0ecd05 100644 --- a/docs/jit-thoughts +++ b/docs/jit-thoughts @@ -11,48 +11,6 @@ We are designing a JIT compiler, so we have to consider two things: The current approach is to keep the JITer as simple as possible, and thus as fast as possible. The generated code quality will suffer from that. -Register allocation is first done inside the trees of the forest, and each -tree can use the full set of registers. We simply split a tree if we get out of -registers, for example the following tree: - - - add(R0) - / \ - / \ - a(R0) add(R1) - / \ - / \ - b(R1) add(R2) - / \ - / \ - c(R2) b(R3) - -can be transformed to: - - - stloc(t1) add(R0) - | / \ - | / \ - add(R0) a(R0) add(R1) - / \ / \ - / \ / \ - c(R0) b(R1) b(R1) t1(R2) - - -Please notice that the split trees use less registers than the original -tree. - -Triggering JIT compilation: -=========================== - -The current approach is to call functions indirectly. The address to call is -stored in the MonoMethod structure. For each method we create a trampoline -function. When called, this function does the JIT compilation and replaces the -trampoline with the compiled method address. - -We should consider using the CACAO approach, they do not use a trampoline at -all. - Register Allocation: ==================== @@ -61,15 +19,6 @@ allocation. For example this is needed by call, which return the value always in EAX on x86. The current implementation works without such system, due to special forest generation. -X86 Register Allocation: -======================== - -We can use 8bit or 16bit registers on the x86. If we use that feature we have -more registers to allocate, which maybe prevents some register spills. We -currently ignore that ability and always allocate 32 bit registers, because I -think we would gain very little from that optimisation and it would complicate -the code. - Different Register Sets: ======================== @@ -86,11 +35,10 @@ call class methods for ALU operations like add or sub. Sure, this method will be be a bit inefficient. The more performant solution is to allocate two 32bit registers for each 64bit -value. We add a new non terminal to the monoburg grammar called long_reg. The +value. We add a new non terminal to the monoburg grammar called "lreg". The register allocation routines takes care of this non terminal and allocates two 32 bit registers for them. - Forest generation: ================== @@ -157,14 +105,6 @@ STLOC(ADD (LDLOC, LDLOC)) This is what lcc is doing, if I understood 12.8, page 342, 343? -Value Types: -============ - -The only CLI instructions which can handle value types are loads and stores, -either to local variable, to the stack or to array elements. Value types with a -size smaller than sizeof(int) are handled like any other basic type. For other -value types we load the base address and emit block copies to store them. - Possible Optimisations: ======================= @@ -180,3 +120,6 @@ else for (i = 0; i < N; i++) { check_range (a, i); a [i] = X; } The "else" is only to keep original semantics (exception handling). +We need loop detection logic in order to implement this (dominator tree). + +AFAIK CACAO also implements this.
\ No newline at end of file |