- 164f927 IR level for support of 128 integer SIMD operations. Use this to do by sewardj · 20 years ago
- 008754b x86 guest: finish SSE2 floating point insns. by sewardj · 20 years ago
- fd22645 x86 guest: Implement a whole bunch of SSE2 instructions, mostly by sewardj · 20 years ago
- 636ad76 Copy-n-paste 32x4 floating point stuff into 64x2 floating point stuff so by sewardj · 20 years ago
- c1e7dfc Finish SSE1 instructions! Finallyatlast. by sewardj · 20 years ago
- 129b3d9 Fix a load of confusion with SSE scalar float insns and memory. by sewardj · 20 years ago
- 0bd7ce6 Even more SSE insns. by sewardj · 20 years ago
- 3bca906 Rationalisation/cleanup of float to/from int conversions and rounding by sewardj · 20 years ago
- 9636b44 x86 guest/host: a whole bunch more SSE instructions. by sewardj · 20 years ago
- 176a59c Add a bunch of easy SSE insns. by sewardj · 20 years ago
- 4cb918d Mucho messing around with x86 FP/SSE rounding modes etc. As a result by sewardj · 20 years ago
- 1e6ad74 x86 guest/host: do SSE comparisons. by sewardj · 20 years ago
- c9a4366 Make a start on SSE for x86 guest. by sewardj · 20 years ago
- 893aada Create a new mechanism: "emulation warnings", which is a way for Vex by sewardj · 20 years ago
- 49bfe67 Properly sanity check IRConst values. by sewardj · 20 years ago
- 0276d4b Rename the boolean field in IRStmt_Exit from 'cond' to 'guard' to be by sewardj · 20 years ago
- ba99931 Rename all the _Bit types to I1 so as to be consistent with all the by sewardj · 20 years ago
- 92d168d Rename INVALID_IRTEMP to IRTemp_INVALID to be consistent with by sewardj · 20 years ago
- 86b133b Get rid of some unused primops: And1 Or1 Neg8/16/32/64. by sewardj · 20 years ago
- f8ed9d8 Add copyright notices. by sewardj · 20 years ago
- fd33277 More primop support needed by memcheck. by sewardj · 20 years ago
- f32c67d Add mkIRExprVec_5. by sewardj · 20 years ago
- 43c5646 IRCallee: add a mcx_mask field, through which front ends can pass by sewardj · 20 years ago
- 8eda630 Add new primop 1Sto16. by sewardj · 20 years ago
- b5874aa Add a couple more primops needed by Memcheck. by sewardj · 20 years ago
- 415d935 Fill in a few more Memcheck-supporting cases. by sewardj · 20 years ago
- 4b861de Make the x86 back end capable of generating conditional calls. by sewardj · 20 years ago
- cf78790 Changes to support Memcheck: by sewardj · 20 years ago
- b8385d8 Give dirty helper calls a guard field (IRExpr* of type Ity_Bit) which by sewardj · 20 years ago
- eeac841 Rename some types and data fields in a more logical way. No change of by sewardj · 20 years ago
- c13e2ed Enhance sanity checking pertaining to IRCallees -- make sure by sewardj · 20 years ago
- f965526 Modified/new convenience functions for constructing helper calls. by sewardj · 20 years ago
- 7735254 Implement regparm (1 to 3) in the x86 back end. by sewardj · 20 years ago
- 8ea867b Instead of denoting helper call targets by their names, add a new type by sewardj · 20 years ago
- 49651f4 Various changes needed to get Addrcheck supported: by sewardj · 20 years ago
- c5fc7aa Changes pertaining to supporting instrumentation: by sewardj · 20 years ago
- 442d0be Implement IEEE754-compliant fprem1. by sewardj · 20 years ago
- 99016a7 Implement guest x86 fptan. by sewardj · 20 years ago
- 6e797c5 Introduce primop Iop_Not1 :: Bit -> Bit, and use it to to a better by sewardj · 20 years ago
- 695cff9 Add loop unrolling to iropt (simple, crude first attempt). This also by sewardj · 20 years ago
- e98dcf2 First attempt at PutI -> GetI forwarding. by sewardj · 20 years ago
- 0bfea7f Small change to printing of guest-state-array descriptors (IRArrays). by sewardj · 20 years ago
- 6d07636 Change the name of some IR fields from .expr to .data, as .expr by sewardj · 20 years ago
- 2d3f77c Major revamp of GetI/PutI in preparation for PutI/GetI forwarding. by sewardj · 20 years ago
- 4345f7a Add a simple common-subexpression removal pass to iropt. Potentially by sewardj · 20 years ago
- edeb4c4 Do a bit of inlining. by sewardj · 20 years ago
- 17442fe Fully install generic IR machinery for dirty helper calls and make it by sewardj · 20 years ago
- 4734104 x86 -> IR -> x86: fix enough bits to make OOo 1.1.2 work. by sewardj · 20 years ago
- 06c32a0 Implement yet more x86 FP transcendental instructions. by sewardj · 20 years ago
- 8308aad Yet more x86 FP cases. by sewardj · 20 years ago
- e5427e8 x86: handle divw/idivw. Needs proper testing. by sewardj · 20 years ago
- 46de407 x86: implement FPREM (sigh, nasty, complicated) by sewardj · 20 years ago
- 52ace3e * Clean up primop naming a bit * Implement x86 FYL2X insn by sewardj · 20 years ago
- 883b00b Fill in many x86 integer and FP cases, enough to get konqueror by sewardj · 20 years ago
- e670911 Implement x86 FRNDINT, taking into account of course the current rounding mode. by sewardj · 20 years ago
- c4be80c * x86 guest: instead of just tracking the FPU's C3, C2 and C0 flags, by sewardj · 20 years ago
- cfded9a Fill in many more x86 FP cases. by sewardj · 20 years ago
- bdc7d21 Make floating-point comparisons work, and fill in a bunch of other x86 by sewardj · 20 years ago
- 8f3debf x86: make float to integer conversions observe the rounding mode in effect. by sewardj · 20 years ago
- 89cd093 Fix a large number of cases at both ends of the pipeline for x86 by sewardj · 20 years ago
- ce646f2 Do a bit more x86 floating point. As a result, had to do 'bsf' and 'bsr' by sewardj · 20 years ago
- 207557a Front end stuff to support floating point on x86 (guest). by sewardj · 20 years ago
- 84ff065 by sewardj · 20 years ago
- b8e7586 Allow constant folding to apply to IRStmt_Exit. by sewardj · 20 years ago
- 39e3f24 Teach iropt to do dead code elimination. by sewardj · 20 years ago
- d7cb853 iropt: implement flattening (into SSA form). This caused various by sewardj · 20 years ago
- a58ea66 More FPU stuff. by sewardj · 20 years ago
- bb53f8c More x86 FP stuff, including conversion stuff to/from Ints. by sewardj · 20 years ago
- d1725d1 Lots of spadework for getting x86 floating point to work. by sewardj · 20 years ago
- b81f8b3 Finally finish pissing around with x86 integer multiply instructions. by sewardj · 20 years ago
- c22a6fd Change the x86 front end's implementation of shift instructions so by sewardj · 20 years ago
- 8c7f1ab Fill in many more cases thrown up the QEMU test program. by sewardj · 20 years ago
- a238471 Fill in / fix many x86->IR->x86 cases exposed by QEMU's test-i386 program. by sewardj · 20 years ago
- 0255207 Fill in misc cases. by sewardj · 20 years ago
- c0ee2ed Make compilation work again after renaming files. by sewardj · 20 years ago[Renamed (99%) from priv/ir/ir_defs.c]
- 443cd9d Loads more x86 insn selector hacking. by sewardj · 20 years ago
- 4042c7e Mucho instruction selector hacking. by sewardj · 20 years ago
- 66de227 Get rid of add with carry and subtract with borrow from the IR. by sewardj · 20 years ago
- eeb9ef8 Get rid of guards on IR Puts, and instead have an IR Mux expression. by sewardj · 20 years ago
- 6d2638e Further improvements to the IR sanity checker. by sewardj · 20 years ago
- 6efd4a1 Complete the IR sanity checker -- at least up to the point where by sewardj · 20 years ago
- 3543921 First part of IR sanity checking: def/use checks. by sewardj · 20 years ago
- e539a40 Rationalise IR a bit: by sewardj · 20 years ago
- 9690d92 x86toIR: reorganise multiply stuff, and make a start on division stuff by sewardj · 20 years ago
- cf780b4 x86toIR: some multiply stuff by sewardj · 20 years ago
- 78c19df Rationalise IR-level control flow: by sewardj · 20 years ago
- 64e1d65 x86toIR: rep-prefixed insns by sewardj · 20 years ago
- 77b86be Yet more x86toIR cases, including set<cond> insn. by sewardj · 20 years ago
- 9334b0f Yet more cases in x86toIR. by sewardj · 20 years ago
- e90ad6a x86toIR: machinery for setting condition codes from shift insns. by sewardj · 20 years ago
- e87b484 More x86 -> IR stuff, including baseline %eflags handling. by sewardj · 20 years ago
- e05c42c A whole bunch more x86 -> IR hacking. by sewardj · 20 years ago
- 41f43bc - Mucho x86 to IR hacking. by sewardj · 20 years ago
- fbcaf33 More stuff. by sewardj · 20 years ago
- 279b88f Print PUT in uppercase. by sewardj · 20 years ago
- c9a6570 First pass at x86 to IR conversion. Compiles, but nothing works. by sewardj · 20 years ago
- 887a11a Rename everything to use the "vex" name. by sewardj · 20 years ago
- 35421a3 Major hashing around to restructure the world. by sewardj · 20 years ago
- ac9af02 Mash around top-level library structure some more. by sewardj · 20 years ago
- 3e1b2d2 Start to move files into a new structure. by sewardj · 20 years ago[Renamed from ir_defs.c]