paweld/fpc
1
0
Fork 0
mirror of https://gitlab.com/freepascal.org/fpc/source.git synced 2025-11-03 10:29:37 +01:00
Code Issues Packages Projects Releases Wiki Activity
49,623 Commits 102 Branches 76 Tags 294 MiB
3f072b3e29
Commit Graph

25 Commits

Author SHA1 Message Date
Jonas Maebe
8c006f04e2 * don't check type size equality for conversions to formaldef either 
git-svn-id: trunk@33987 -
 2016-06-15 18:31:37 +00:00
Jonas Maebe
3f01efbee0 * adapted tcgtypeconvnode.second_int_to_int() to work with high level cpu 
targets, and removed the (buggy) LLVM-specific version

git-svn-id: trunk@33980 -
 2016-06-12 20:56:40 +00:00
Jonas Maebe
2d1023140a * don't try to check whether the size of from and to in type conversion nodes 
are the same in case of open arrays (their size can't be determined at
    compile time) -- this is just a sanity check, which can't be performed in
    this case

git-svn-id: trunk@33979 -
 2016-06-12 20:56:37 +00:00
Jonas Maebe
9d2bba1917 * create a separate type and def for the LLVM "i1" type, because reusing 
pasbool8type for this results in too much trouble (we mustn't use i1
    for parameters, because then LLVM will try to apply the ABI convention
    for passing "1 bit" values, or in records because then this may
    result in unwanted bitpacking). Downside: the new LLVMBool1 type is
    also exposed in the system unit, because we need it to define LLVM
    intrinsics...

git-svn-id: trunk@33726 -
 2016-05-20 20:51:44 +00:00
Jonas Maebe
49a83b2872 * allow targets to keep certain type conversions for equal types in 
inserttypeconv()
   o keep typeconversions between structurally equivalent but semantically
     different procvardefs for LLVM and JVM, because they're different
     types there

git-svn-id: trunk@32904 -
 2016-01-10 14:01:46 +00:00
Jonas Maebe
e3d2980c98 * don't load data in a register when converting from void to an integer 
type (like with integer(pointer1^):=integer(pointer2^))

git-svn-id: trunk@32743 -
 2015-12-26 20:01:44 +00:00
Jonas Maebe
a9498b6735 * fixed the parameter order for thlcgobj.g_load_const_field_by_name() 
git-svn-id: trunk@32595 -
 2015-12-05 18:03:48 +00:00
Jonas Maebe
df9efdeb4c * fixed llvm types for loading the address of a nested routine 
git-svn-id: trunk@32594 -
 2015-12-05 18:03:45 +00:00
Jonas Maebe
35714e5d50 * don't internallerror when encountering an "absolute" type conversion to 
a differently sized def in LLVM

git-svn-id: trunk@32571 -
 2015-12-02 22:29:23 +00:00
Jonas Maebe
13cf405f8c * fixed llvm second_int_to_real for comp 
git-svn-id: trunk@32464 -
 2015-11-22 11:49:11 +00:00
Jonas Maebe
72c7cb2770 * typecast second_proc_to_procvar() results for llvm if the generic code 
performed a simple location_copy() from one complex procvar type to
    another

git-svn-id: trunk@32426 -
 2015-11-21 12:37:50 +00:00
Jonas Maebe
423eb7ddb6 * fixed int_to_real() conversion llvm destination type when the result is 
currency

git-svn-id: trunk@32423 -
 2015-11-21 12:37:40 +00:00
Jonas Maebe
bd4787c716 * force complex procvars immediately to memory for llvm, so we don't have to 
deal with the hacky register/registerhi complex locations

git-svn-id: trunk@32407 -
 2015-11-21 12:36:48 +00:00
Jonas Maebe
4f54f8b3bd * fixed expectloc for llvm bool_2_int typeconversions where the location 
doesn't change

git-svn-id: trunk@32248 -
 2015-11-04 22:23:50 +00:00
Jonas Maebe
0fc1fd6ac1 * replaced current_procinfo.currtrue/falselabel with storing the true/false 
labels of LOC_JUMP in the node's location. This generates some extra jumps
    for short circuit boolean and/or-expressions if optimizations are off, but
    with optimisations enabled the generated code is the same (except for JVM
    because the jump threading optimisation isn't enabled there yet).

git-svn-id: trunk@31431 -
 2015-08-27 18:28:57 +00:00
Jonas Maebe
f402b0d7df * changed getpointerdef() into a tpointerdef.getreusable() class method 
o allows removing the ugly x86 hacks

git-svn-id: trunk@31144 -
 2015-06-22 08:17:49 +00:00
Jonas Maebe
eae68e86a1 * don't insert a typeconversion from pasbool to int in case the left 
location was LOC_JUMP or LOC_FLAGS, as in that case the generic
    code has already done this using location_force_reg()

git-svn-id: trunk@30783 -
 2015-05-03 16:51:10 +00:00
Jonas Maebe
f7b5ee64ae * always insert type conversion for pasbool to int on llvm, as pasbool is 
represented by i1 while equally sized integers are i8

git-svn-id: trunk@30709 -
 2015-04-25 15:50:58 +00:00
Jonas Maebe
0aab46b97f + tllvmtypeconvnode.second_int_to_bool() implementation 
git-svn-id: branches/hlcgllvm@28370 -
 2014-08-10 19:40:13 +00:00
Jonas Maebe
9fada1a238 * don't internalerror when typecasting voidpointer^ or formaldef 
git-svn-id: branches/hlcgllvm@28208 -
 2014-07-12 22:42:15 +00:00
Jonas Maebe
27e350fa9a + llvm implementation of second_int_to_real() 
git-svn-id: branches/hlcgllvm@28207 -
 2014-07-12 22:42:11 +00:00
Jonas Maebe
4746e17be5 * implemented second_int_to_int() for llvm 
o the generic implementation ignores the need for bitcasts from pointers to
     ordinals and vice versa, and also hacks around with offsets

git-svn-id: branches/hlcgllvm@28206 -
 2014-07-12 22:14:37 +00:00
Jonas Maebe
b494f76b81 * made internalerror unique 
git-svn-id: branches/hlcgllvm@28201 -
 2014-07-12 22:13:39 +00:00
Jonas Maebe
cdfb23bf6c * converted some leftovers in tcgtypeconvnode.second_int_to_int/ 
second_real_to_real to thlcgobj
  + added type conversion to second_pointer_to_array for llvm

git-svn-id: branches/hlcgllvm@27000 -
 2014-03-06 21:41:27 +00:00
Jonas Maebe
8cc9b64ad7 * don't discard typeconversion nodes that only change the result type when 
compiling for llvm, because this type change also needs to be modelled in
    llvm IR

git-svn-id: branches/hlcgllvm@26999 -
 2014-03-06 21:41:24 +00:00