florian c75486db89 * patch by Nico Erfurth: ... Reorder unaligned Load sequence on ARM The old version produced code like that: ldrb rDEST, [rBASE] ldrb rTemp, [rBASE, #1] orr rDEST, rDEST, rTEMP lsl #8 (2 stall cycles) ldrb rTemp, [rBASE, #2] orr rDEST, rDEST, rTEMP lsl #16 (2 stall cycles) ldrb rTemp, [rBASE, #3] orr rDEST, rDEST, rTEMP lsl #24 (2 stall cycles) This creates a lot of stall-cycles on ARM Implementations with load delay slots like Marvel Kirkwood or Intel XScale. With the usual up to 2 stall-cycles this code requires a total of 13 cycles (7 instructions + 6 stall cycles) in best case. The new code uses a second temp register to avoid the stall cycles. ldrb rDEST, [rBASE] ldrb rTemp1, [rBASE, #1] ldrb rTemp2, [rBASE, #2] orr rDEST, rDEST, rTEMP1 lsl #8 ldrb rTemp1, [rBASE, #3] orr rDEST, rDEST, rTEMP2 lsl #16 orr rDEST, rDEST, rTEMP1 lsl #24 (1 stall cycle) The rescheduling and second register bring the total cycles down to 8. If a later rescheduling should happen for the last orr it even can go down to 7. git-svn-id: trunk@21363 -		2012-05-22 19:09:20 +00:00
..
aasmcpu.pas	* skip comments properly when searching for places for constant pool distances	2012-05-15 18:08:19 +00:00
agarmgas.pas	+ generic implementation of ReplaceForbiddenAsmSymbolChars() instead	2012-04-11 18:01:57 +00:00
aoptcpu.pas	* patch by Nico Erfurth:	2012-05-17 08:31:44 +00:00
aoptcpub.pas	* typo fixed	2012-03-11 08:24:44 +00:00
aoptcpuc.pas
aoptcpud.pas
armatt.inc	+ patch by Bernd to add the push/pop mnemonic for arm/thumb-2, resolves #22041	2012-05-15 18:52:09 +00:00
armatts.inc	+ patch by Bernd to add the push/pop mnemonic for arm/thumb-2, resolves #22041	2012-05-15 18:52:09 +00:00
armins.dat	+ patch by Bernd to add the push/pop mnemonic for arm/thumb-2, resolves #22041	2012-05-15 18:52:09 +00:00
armnop.inc	+ support for nop, msr and mrs instructions	2009-01-26 14:18:42 +00:00
armop.inc	+ patch by Bernd to add the push/pop mnemonic for arm/thumb-2, resolves #22041	2012-05-15 18:52:09 +00:00
armreg.dat	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
armtab.inc	+ support for nop, msr and mrs instructions	2009-01-26 14:18:42 +00:00
cgcpu.pas	* patch by Nico Erfurth:	2012-05-22 19:09:20 +00:00
cpubase.pas	* patch by Nico Erfurth:	2012-05-17 08:03:51 +00:00
cpuinfo.pas	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
cpunode.pas	* the objc1 unit has been renamed to objc	2009-09-27 15:24:50 +00:00
cpupara.pas	* use correct result registers for in64 results on armbe, resolves #21731	2012-04-20 18:07:06 +00:00
cpupi.pas	* use r7 instead of r11 as frame pointer on Darwin/iOS, and make sure r7	2012-03-29 20:54:33 +00:00
cputarg.pas	+ some generic changes preparing for darwin/arm support	2008-10-02 15:10:13 +00:00
hlcgcpu.pas	* create/destroy also the high level code generator for all architectures,	2011-08-20 07:21:16 +00:00
itcpugas.pas
narmadd.pas	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
narmcal.pas	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
narmcnv.pas	* moved subsetref/reg and bit_set/test support from cgobj to hlcgobj for	2012-05-13 12:33:10 +00:00
narmcon.pas	* fixed ARM and MIPS compilation after r14912	2010-02-18 21:19:17 +00:00
narminl.pas	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
narmmat.pas	* moved subsetref/reg and bit_set/test support from cgobj to hlcgobj for	2012-05-13 12:33:10 +00:00
narmset.pas	* converted tcgcasenode.pass_generate_code() to hlcgobj	2011-08-20 07:48:33 +00:00
pp.lpi.template	* improved template with help from Mattias Gaertner	2006-08-28 20:29:04 +00:00
raarm.pas	o patch by Jeppe Johansen to fix mantis #17472:	2010-12-24 15:54:39 +00:00
raarmgas.pas	+ support for REV and several other ARMv6/ARMv6T2+ opcodes (mantis #21888)	2012-05-13 12:14:26 +00:00
rarmcon.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rarmdwa.inc	o added ARM VPFv2/VFPv3 support:	2009-12-03 22:46:30 +00:00
rarmnor.inc	o added ARM VPFv2/VFPv3 support:	2009-12-03 22:46:30 +00:00
rarmnum.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rarmrni.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rarmsri.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rarmsta.inc	o added ARM VPFv2/VFPv3 support:	2009-12-03 22:46:30 +00:00
rarmstd.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rarmsup.inc	+ support for the ARM hard float EABI on Linux (patch by Peter Green):	2012-03-29 20:50:09 +00:00
rgcpu.pas	* patch by Jeppe Johansen to avoid corruption of frame/stack pointer by pre/post indexed operations, resolves #19679	2011-08-16 22:43:30 +00:00