paweld/fpc
1
0
Fork 0
mirror of https://gitlab.com/freepascal.org/fpc/source.git synced 2025-04-27 08:33:43 +02:00
Code Issues Packages Projects Releases Wiki Activity
60,068 Commits 100 Branches 75 Tags 297 MiB
d1cd26592a
Commit Graph

313 Commits

Author SHA1 Message Date
florian
b36597c76a * better calculation of estimated stack frame size 
git-svn-id: trunk@46733 -
 2020-08-30 20:53:59 +00:00
yury
7bffafdf8b * Improved the parentfp optimization to properly handle cases when a nested routine calls other nested routines. 
git-svn-id: trunk@45665 -
 2020-06-20 18:34:26 +00:00
yury
c15b6f4516 * Fixed the parentfp optimization for some cases when nested procvars are used. 
git-svn-id: trunk@45664 -
 2020-06-20 18:28:07 +00:00
yury
2808873d1b * Reworked the optimization of unused $parentfp for nested routines. 
- Do not remove the $parentfp parameter as was done in the previous optimization approach. Instead when $parentfp is unused to the following:
      - On the caller side: Omit passing the value for $parentfp for targets where tcgcallparanode.push_zero_sized_value_para=false (classic CPU targets). 
          Pass 0/nil as $parentfp for targets where tcgcallparanode.push_zero_sized_value_para=true;
      - On the callee side: Prevent allocation of registers/resources for $parentfp.
  - When possible keep $parentfp in a register.
  - Set the pio_nested_access flag in tprocinfo.set_needs_parentfp() to properly handle deep nesting levels;

git-svn-id: trunk@45436 -
 2020-05-19 13:17:47 +00:00
florian
fc98a0db4f * cosmetics 
git-svn-id: trunk@44542 -
 2020-04-03 20:15:25 +00:00
florian
fa4cbc89a5 + Xtensa: hard float support, i.e. make use of floating point extension if available 
git-svn-id: trunk@44539 -
 2020-04-03 20:15:23 +00:00
florian
f6c16323fa * unified loadnf_load_self_pointer into loadnf_load_addr 
+ var parameters are now allowed when doing tail recursion optimziation, resolves #32811

git-svn-id: trunk@43824 -
 2019-12-30 22:43:10 +00:00
Jonas Maebe
ac1e0f96bd * replaced tentryfile.get/putsmall/normalset() with a common tget/putset 
that expects an open array of byte, and use it for all sets
   o since all sets need to be typecasted to an array type of the appropriate
     size, we'll get a compilation error in case this needs to be done and
     that also tells us at the same time that the ppu version will need to
     be increased
  * enabled {$packset 1} for the compiler, as this is now safe with the above
    changes

git-svn-id: trunk@43407 -
 2019-11-06 21:50:19 +00:00
Jonas Maebe
ef6dde6de3 * fixed loading the address of a copied labelnode in a loadnode 
(mantis #35877)

git-svn-id: trunk@42987 -
 2019-09-13 19:06:40 +00:00
pierre
243c967967 Commit of new debug feature implemented by J. Gareth Moreton 
Allows compilation of compiler using -dDEBUG_NODE_XML
  which will generate a NAME-node-dump.xml file for each 
  unit, program or library compiled,
  containing a XML description of the nodes handled during
  compilation of the unit, program or library.

git-svn-id: trunk@42271 -
 2019-06-22 14:08:47 +00:00
yury
e5c2d13671 * Do not set pi_needs_got in current_procinfo.flags at the node level, since the GOT usage can only be estimated there. Instead set the pi_needs_got flag at places where the GOT register is accessed during the code generation. This eliminates generation of the unneeded initialization of the GOT register and fixes linker errors when the _GLOBAL_OFFSET_TABLE_ symbol is referenced but no actual GOT references are present. 
git-svn-id: trunk@41460 -
 2019-02-25 13:35:40 +00:00
Jonas Maebe
51e68eb302 * fixed several places where the interface crc could change: 
o unsetting po_inline while parsing the implementation for various reasons
     (interprocedural goto/label, accessing a local in a parent frame,
      having nested procedures)
   o instead handle this via the pio_inline_not_possible flag
   o noreturn can no longer be specified only in the implementation

git-svn-id: trunk@40789 -
 2019-01-06 20:35:56 +00:00
florian
ae8304f657 * set nf_write flag properly for vec. nodes of types with implicit dereferencing 
git-svn-id: trunk@40339 -
 2018-11-17 15:08:23 +00:00
florian
9f16c34329 + initial work for tls-based threadvar support on arm-linux 
git-svn-id: trunk@40267 -
 2018-11-07 22:02:58 +00:00
svenbarth
a8b2ad31d1 * fix for Mantis #34355: correctly check whether the array constructor is empty (aka []) or not 
git-svn-id: trunk@39882 -
 2018-10-07 12:25:24 +00:00
svenbarth
85439a0fa0 * also check for nf_assign_done_in_right flag after typechecking the right side 
git-svn-id: trunk@39118 -
 2018-05-25 15:54:41 +00:00
florian
0fd0e356e1 + new flag ra_different_scope: used if a node tree is marked as non-regable if a variable is accessed from a different scope 
git-svn-id: trunk@38689 -
 2018-04-04 21:39:47 +00:00
florian
c230f81719 + support for <dyn. array>+<dyn. array>, resolves #30463 
git-svn-id: trunk@38406 -
 2018-03-04 16:12:43 +00:00
svenbarth
d50848174a + add putboolean and getboolean convenience methods to tentfile 
* use putboolean and getboolean where approbiate

git-svn-id: trunk@37972 -
 2018-01-14 21:36:02 +00:00
nickysn
efc5e339d0 * use an enum instead of integer constants to represent inline numbers 
* compinnr.inc include file converted to a unit
* inline number field size stored in ppu increased from byte to longint
* inlines in the parse tree (when written with the -vp option) now printed with
  their enum name, instead of number

git-svn-id: trunk@36174 -
 2017-05-10 14:41:43 +00:00
florian
b1dff29cbf * removed unused units 
git-svn-id: trunk@36165 -
 2017-05-09 19:53:14 +00:00
svenbarth
8f4ef91efb * tarrayconstructornode: force conversion to set if there's a range node in the tree 
git-svn-id: trunk@36101 -
 2017-05-04 21:55:12 +00:00
svenbarth
82b2d22ed6 * keep track of allow_array_constructor as part of the array constructor node 
git-svn-id: trunk@36099 -
 2017-05-04 21:52:21 +00:00
svenbarth
dfc23da06c * tassignmentnode.pass_typecheck: also allow "dynarray := []" in addition to "dynarray := nil" 
git-svn-id: trunk@36098 -
 2017-05-04 21:40:37 +00:00
svenbarth
b1c66dee9b + tarrayconstructornode: add utility method has_range_node to check whether the array constructor node tree contains any range node 
git-svn-id: trunk@36091 -
 2017-05-04 21:16:47 +00:00
svenbarth
4a2266ac44 * tarrayconstructornode.pass_typecheck: set elementdef of the arraydef only after the arrayoptions are changed 
git-svn-id: trunk@36090 -
 2017-05-04 21:03:33 +00:00
nickysn
012153201a * moved the conversion of "x:=x op k" to inline nodes to a separate optimization 
pass, so that it can be done after other optimizations, such as constant
  propagation

git-svn-id: trunk@35961 -
 2017-04-26 22:44:01 +00:00
nickysn
5bb77905cd * fixed the missing warnings for unused variables at -O3, when one of the new 
modify-in-place inline nodes are used

git-svn-id: trunk@35773 -
 2017-04-11 14:25:57 +00:00
nickysn
4697a653a4 + perform the "i:=i +/-/and/or/xor k" optimization when there are two typecasts 
inserted. This makes the optimization work for more integer types.

git-svn-id: trunk@35772 -
 2017-04-11 11:25:40 +00:00
nickysn
d8406c4227 + perform the "i:=-i" / "i:=not i" optimization even when there are typecasts 
inserted; this makes the optimization work for all integer types

git-svn-id: trunk@35763 -
 2017-04-10 14:06:19 +00:00
nickysn
98be270adb * fixed a memory leak in the "i:=-i" / "i:=not i" optimization 
git-svn-id: trunk@35762 -
 2017-04-10 13:27:29 +00:00
nickysn
e78f0aafcd - remove whitespace, was inserted by accident 
git-svn-id: trunk@35760 -
 2017-04-10 11:55:26 +00:00
nickysn
3616627d0b * at -O3 optimization level, convert "i:=not i" and "i:=-i" to the new 
in_not_assign_x and in_neg_assign_x inline nodes, which may generate better
  code on certain cpus (such as x86)

git-svn-id: trunk@35758 -
 2017-04-09 14:55:15 +00:00
nickysn
f29492bdea * at -O3 level, convert i:=i and/or/xor k to the new in_[and/or/xor]_assign_x_y 
inline nodes, which should generate better code for certain CPU targets,
  including x86. Note that the optimization isn't applied yet for all integer
  types (those that have extra implicit typecasts, inserted by the compiler,
  aren't handled yet).

git-svn-id: trunk@35685 -
 2017-03-29 15:20:57 +00:00
florian
b55cb246da * do +/-/pred/succ to inc/dec optimizations only at an appropriate optimziation level 
git-svn-id: trunk@35569 -
 2017-03-12 07:49:37 +00:00
pierre
a7c7319352 Avoid use of getcopy method for i:=i+/i k optimization 
git-svn-id: trunk@35568 -
 2017-03-11 21:41:51 +00:00
florian
a679cd4e37 * copy loadnodeflags in tloadnode.dogetcopy as well 
git-svn-id: trunk@35558 -
 2017-03-10 20:44:20 +00:00
nickysn
bd48a71a98 * replace i:=i+k/i:=k+i/i:=i-k (integer only, for now; later we can also do this for pointers too) by inc/dec(i,k) 
git-svn-id: trunk@35550 -
 2017-03-09 15:35:27 +00:00
florian
a27b07b342 + more restrictions on pred/succ to dec/inc optimization 
o check if the argument can be used as a call by reference parameter
  o check if the argument has no side-effects
+ tests

git-svn-id: trunk@34816 -
 2016-11-06 16:36:02 +00:00
florian
000cffa8e3 * make more use of nf_internal to avoid range check errors 
git-svn-id: trunk@34793 -
 2016-11-05 23:05:51 +00:00
florian
509f6839fe + replace i:=succ/pred(i) by inc/dec(i) 
git-svn-id: trunk@34791 -
 2016-11-05 23:05:48 +00:00
Jonas Maebe
233622157c * don't attempt to load the VMT of a niln when taking the address of a 
class method (mantis #30706)

git-svn-id: trunk@34641 -
 2016-10-07 16:41:32 +00:00
Jonas Maebe
2d051f89f7 * don't increase refcounts for variants assigned to (ti_)const nodes, fixes 
memory leak after r34288 and should have been done as part of r34287
    (mantis #30546)

git-svn-id: trunk@34440 -
 2016-09-06 21:28:34 +00:00
Jonas Maebe
f64556c125 * when taking the address of a class method via an instance, create a procvar 
with the VMT of the instance as self instead of the self instance pointer
    (mantis #29491)

git-svn-id: trunk@34395 -
 2016-08-30 07:25:16 +00:00
Jonas Maebe
18580d3069 * when assigning a value to a temp marked as ti_const, do not immediately 
release the source location in case it was also a temp. Reason: we
    don't increase reference counts when assigning to a ti_const (they are
    like const parameters), so the original value must stay alive until
    the ti_const temp is freed
   o free the original data in the ttempdelete node for the ti_const temp
   o don't increase reference counts when assigning composite types to
     ti_const temps either

git-svn-id: trunk@34287 -
 2016-08-12 13:35:51 +00:00
Jonas Maebe
9914f02faa * always set the resultdef for variant array constructors correctly, so we 
don't need to pull all kinds of special tricks pass_generate_code

git-svn-id: trunk@32233 -
 2015-11-04 20:46:10 +00:00
Jonas Maebe
95927665ce - removed variable notification support, it's not used and superceded by 
DFA

git-svn-id: trunk@31916 -
 2015-10-02 17:00:25 +00:00
florian
63f828d54e * take care of nf_internal when doing range checking and inserting type conversions in tassignmentnode.pass_typecheck 
git-svn-id: trunk@31473 -
 2015-08-31 20:03:54 +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
dd3d0ae5bc * set the type of the hidden $vmt parameter to the correct classrefdef and 
adjust the parameter comparison code to ignore types of $vmt parameters
    when comparing (previously they were all voidpointer and hence also
    equal)

git-svn-id: trunk@30955 -
 2015-05-31 16:51:02 +00:00