Commit Graph

185 Commits

Author SHA1 Message Date
Jonas Maebe
1a559f5c60 * move the homogenous float aggregate determination function to a common
unit for ARM and AArch64

git-svn-id: trunk@41419 -
2019-02-23 15:42:39 +00:00
yury
382d5060a0 * aarch64: According to ARM64 ABI, function's arguments and the result are not required to be sign/zero extended to a full register size. Obey this rule for all targets except iOS where sign/zero extension is required.
git-svn-id: trunk@41412 -
2019-02-22 17:59:54 +00:00
Jonas Maebe
ac8411f3cf * pass dynamic array parameters of cdecl routines by value on AArch64
(like on onther platforms)
   o it was passed by reference as a side effect of the fact arrays in C are
     pointers, but a dynamic array already is a pointer

git-svn-id: trunk@40573 -
2018-12-16 20:44:35 +00:00
Jonas Maebe
95171a1e22 * initialise cgpara with valid data for the C-style "array of const" para
(necessary for the LLVM backend)

git-svn-id: trunk@40571 -
2018-12-16 20:44:28 +00:00
Jonas Maebe
820d2f7135 * support OS_32/OS_64 in AArch64 cgsize2subreg() for MM registers (can happen
for records with floating point fields with the LLVM code generator)

git-svn-id: trunk@40570 -
2018-12-16 20:44:24 +00:00
Jonas Maebe
ff002e2664 * set paralocations correctly for LLVM/AArch (must be register-type instead
of parameter-type based)

git-svn-id: trunk@40568 -
2018-12-16 20:44:18 +00:00
Jonas Maebe
6b58185bda * fixed shifval for big endian (bits instead of bytes)
git-svn-id: trunk@40567 -
2018-12-16 20:44:15 +00:00
Jonas Maebe
6e453a1477 * fixed paraloc def when changing the size to OS_32
git-svn-id: trunk@40566 -
2018-12-16 20:44:12 +00:00
Jonas Maebe
52fec8a94a * when optimising subsetreg moves for aarch64, take into account the fact
that the subsetreg itself can be 32 or 64 bit (mantis #33607)

git-svn-id: trunk@40512 -
2018-12-09 14:46:52 +00:00
florian
9b0ff05ee8 - get rid of MaxOps, it is redundant with max_operands
* MatchOpType with three operands is only available of max_operands>2

git-svn-id: trunk@40190 -
2018-11-02 21:32:29 +00:00
yury
4357caaad8 * Removed unused local vars.
git-svn-id: trunk@40183 -
2018-11-02 18:44:29 +00:00
florian
112fb4613b * take care of the fact that x and w registers overlap when carrying out the Str/LdrAdd/Sub2Str/Ldr Postindex optimization
git-svn-id: trunk@40023 -
2018-10-24 17:38:51 +00:00
florian
0a5e6d29cb + implement assembler optimization Str/LdrAdd/Sub2Str/Ldr Postindex done
git-svn-id: trunk@40017 -
2018-10-23 20:41:27 +00:00
florian
379c2fe821 * cleanup debug code
git-svn-id: trunk@40016 -
2018-10-22 21:27:36 +00:00
florian
f27fee5566 + make use of the BFI instruction in tcgaarch64.a_load_ref_reg_unaligned
git-svn-id: trunk@40015 -
2018-10-22 21:23:48 +00:00
pierre
92acd38f40 Fix for bug report #34380
git-svn-id: trunk@39986 -
2018-10-18 20:21:54 +00:00
yury
4643a0b1ff * Register external gas assembler for aarch64-android and x86_64-android.
git-svn-id: trunk@39969 -
2018-10-18 11:48:27 +00:00
yury
93d672bf16 * aarch64: Fixed loading of a function result which is returned in 4 32-bit MM registers to 2 64-bit int registers. Bug #30329.
git-svn-id: trunk@39953 -
2018-10-17 11:41:34 +00:00
yury
671d31df1d + Added support for the aarch64-android target.
git-svn-id: trunk@39862 -
2018-10-06 09:33:09 +00:00
Jonas Maebe
4686f61002 * keep track of the temp position separately from the offset in references,
so that they can still be freed after the reference has been changed
    (e.g. in case of array indexing or record field accesses) (mantis #33628)

git-svn-id: trunk@38814 -
2018-04-22 17:03:16 +00:00
nickysn
518cdf9674 * replaced the saved_XXX_registers arrays with virtual methods inside
tcpuparamanager, very similar to the existing get_volatile_registers_XXX. The
  new methods are called get_saved_registers_XXX, where XXX is the register
  type ("int", "address", "fpu" or "mm")

git-svn-id: trunk@38794 -
2018-04-19 21:22:16 +00:00
florian
efef9125b5 * another compilation fix
git-svn-id: trunk@38211 -
2018-02-11 18:39:49 +00:00
Jonas Maebe
1b66995754 * factored out check to determine whether a variable can be subscripted in
inline assembly, and fixed check after r35959 (mantis #32318)
   o can also subscript parameters passed by value on the stack
   o can also subscript local variables, the parameters passed by reference
     that are subsequently copied into a local

git-svn-id: trunk@37886 -
2018-01-01 14:29:21 +00:00
florian
a1a14eaaa8 * fix aarch64 building
git-svn-id: trunk@37025 -
2017-08-22 08:44:56 +00:00
Jonas Maebe
61af0fb72d * only take into account the location of the parameter at the callee side to
determine whether it's in a register if it's a pure assembler routine
  * you can't "index" implicit pointers either using their fields

git-svn-id: trunk@36287 -
2017-05-21 20:17:11 +00:00
Jonas Maebe
aa82e00615 * fixed check to determine whether a record parameter can be subscripted
directly in inline assembly: that's only possible if it's a register
    parameter where the address of the record was passed (rather than the
    record itself), or if a parameter has been explicitly typecasted in
    Intel-style assembly using ".size"

git-svn-id: trunk@35959 -
2017-04-26 19:43:35 +00:00
nickysn
c8487c4150 + added individual bits of the x86 flags register as subregisters
git-svn-id: trunk@35955 -
2017-04-26 13:52:52 +00:00
florian
73c46a5988 - removed unused constants
git-svn-id: trunk@35664 -
2017-03-26 13:06:34 +00:00
Jonas Maebe
015f034904 * reverted r35424, wasn't ready for commit yet
git-svn-id: trunk@35426 -
2017-02-11 21:21:44 +00:00
Jonas Maebe
4d9617da97 * fixed check to determine whether a record parameter can be subscripted
directly in inline assembly: that's only possible if it's a register
    parameter where the address of the record was passed (rather than the
    record itself)

git-svn-id: trunk@35424 -
2017-02-11 19:57:08 +00:00
pierre
1b1fe68f92 Fix aarch64 cycle after constant parser simplification change in rev 35229-35230
git-svn-id: trunk@35239 -
2017-01-05 13:41:44 +00:00
Jonas Maebe
880d438704 * renamed t<cpuname>procinfo to tcpuprocinfo for all targets, so we can
inherit from it for LLVM without a thousand ifdefs

git-svn-id: trunk@35141 -
2016-12-16 22:41:21 +00:00
Jonas Maebe
a25ebbba3e + added volatility information to all memory references
o separate information for reading and writing, because e.g. in a
     try-block, only the writes to local variables and parameters are
     volatile (they have to be committed immediately in case the next
     instruction causes an exception)
   o for now, only references to absolute memory addresses are marked
     as volatile
   o the volatily information is (should be) properly maintained throughout
     all code generators for all archictures with this patch
   o no optimizers or other compiler infrastructure uses the volatility
     information yet
   o this functionality is not (yet) exposed at the language level, it
     is only for internal code generator use right now

git-svn-id: trunk@34996 -
2016-11-27 18:17:37 +00:00
Jonas Maebe
86876ed114 * MaybeRecordOffset: initialise "mangledname" variable
git-svn-id: trunk@34857 -
2016-11-09 19:51:36 +00:00
Jonas Maebe
74a49b5f91 * restructured the the TExternalAssembler constructors so that the
hack for the Jasmin descendent is no longer needed

git-svn-id: trunk@34852 -
2016-11-09 19:51:20 +00:00
pierre
89c1b68b14 * Delete regvars unit.
This unit was empty unless OLDREGVARS macro was set,
    but this does not compile and no change has been made since 2011.

  * Remove regvars from all _USES clauses.

git-svn-id: trunk@34808 -
2016-11-06 14:01:39 +00:00
Jonas Maebe
f891788a86 * simplification of and fix for AArch64 cpupara.is_hfa_internal() (patch by
modulo7, mantis #30381)
   o check for array elements of size 0 does not make sense after we've
     established it's an array of floats
   o since the recursive call to is_hfa_internal() already checks for
     floats, we don't have to do it before calling it again for the
     elements of an array
   o set the result to "true" if the array is small enough to be a HFA

git-svn-id: trunk@34293 -
2016-08-12 16:53:02 +00:00
Jonas Maebe
aa1be3276f - removed default value of _typ parameter of TAsmData.(Weak)RefAsmSymbol():
it was AT_NONE, which is invalid and should never be used
  * explicitly pass the correct value for all calls to those methods elsewhere
    in the compiler

git-svn-id: trunk@34250 -
2016-08-05 07:09:16 +00:00
Jonas Maebe
a0efde8167 * automatically generate necessary indirect symbols when a new assembler
symbol is defined
   o removed all places where AB_INDIRECT symbols were explicitly generated
   o only generate AB_INDIRECT symbols for AT_DATA on systems_indirect_var_imports
   o for some symbols an indirect symbol is always required (because they are
     dereferenced by code in RTL units) -> use new AT_DATA_FORCEINDIRECT type

git-svn-id: trunk@34165 -
2016-07-20 20:53:03 +00:00
Jonas Maebe
1cb8c0d00c * specify the def of assembler level symbols defined via
tasmdata.DefineAsmSymbol() and all routines that call it
   o will be used to automatically generate AB_INDIRECT sybols when
     necessary

git-svn-id: trunk@34164 -
2016-07-20 20:52:59 +00:00
Jonas Maebe
8d77de6b22 - removed InsertPData leftover from ARM copy of this file
git-svn-id: trunk@34162 -
2016-07-20 20:52:53 +00:00
Jonas Maebe
0ed6c3d80e * also use the last floating point parameter register on AArch64 in case of
a HFA type (which in practice also applies to plain floating point types)
    (patch by Alfred, mantis #30207)

git-svn-id: trunk@33922 -
2016-06-05 20:18:24 +00:00
Jonas Maebe
1feb062f7d * similar change as in r33393 but now for a_load_reg_ref_unaligned(), although
it's unlikely it would ever cause a problem there

git-svn-id: trunk@33594 -
2016-05-01 12:35:06 +00:00
Jonas Maebe
2a1f2b9fd9 * fixed a_load_regconst_subsetreg_intern() when loading a 32 bit register
to a non-zero bit offset in a subsetreg (mantis #29933)

git-svn-id: trunk@33498 -
2016-04-13 17:09:31 +00:00
Jonas Maebe
8c0d9b581c * don't write the destination register in a_load_ref_reg_unaligned() before
the reference has been used for the last time, as the destination
    register could be the base or index register of the reference

git-svn-id: trunk@33393 -
2016-03-30 20:04:31 +00:00
Jonas Maebe
322493c195 * fixed overflow checking on AArch64 for signed multiplications with zero
(mantis #29912)

git-svn-id: trunk@33380 -
2016-03-29 09:50:42 +00:00
Jonas Maebe
73a303c899 * no need to call maybeadjustresult() after performing a 64 bit operation
git-svn-id: trunk@33379 -
2016-03-29 09:50:39 +00:00
Jonas Maebe
fa3b0ca312 * support marking defs created via the getreusable*() class methods as
"don't free even if not registered"; use for defs that may not be written
    to a ppu file, but that must nevertheless survive the compilation of the
    current module
  * mark all defs created for para locations as "don't free even if not
    registered", because we don't discard and recalculate all para locations
    after a module has been compiled (since that's not needed)
   o solves issues if the paralocations for a routine in the interface of
     unit A are calculated while the implementation of unit B gets
     compiled, and a new reusable type is allocated at that point which
     is not used anywhere else (after r32160)

git-svn-id: trunk@32235 -
2015-11-04 20:46:18 +00:00
Jonas Maebe
7c594b0288 + added support for using Clang as an assembler, and make it the default
for all non-ppc(32/64) Darwin platforms
   o pass the macosx-version-min/iphoneos-version-min to clang as an assembler,
     so that it properly sets this information starting with Xcode 7 (solves
     errors when targeting the iOS simulator, and warnings about object files
     being compiled for a different OS X version when targeting (Mac) OS X)
   o the old assembler is still selectable via -Aas-darwin (required with
     Xcode 3.1.x and older)
   o since the first Xcode version that shipped with Clang is Xcode 3.2, which
     is available for Mac OS X 10.6, most users should not encounter any issues
     with the new default (in fact, it fixes some tests for x86 because Clang
     supports some instructions that "as" doesn't). Clang does not support
     Stabs however, so -gs does require the use of -Aas-darwin

git-svn-id: trunk@31830 -
2015-09-25 18:31:54 +00:00
Jonas Maebe
1131b08d6d * changed idtext of as_darwin to AS-DARWIN, so it can be explicitly
selected via -Aas-darwin (since it compares uppercased strings)

git-svn-id: trunk@31808 -
2015-09-24 06:49:29 +00:00