[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Patch] PVM low-level cleanup and optimization, good version [Was: R
|
From: |
Luca Saiu |
|
Subject: |
Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization] |
|
Date: |
Sun, 01 Dec 2019 14:49:54 +0100 |
|
User-agent: |
Gnus (Gnus v5.13), GNU Emacs 27.0.50, x86_64-pc-linux-gnu |
On 2019-12-01 at 13:40 +0100, Jose E. Marchesi wrote:
> I am waiting for your approval before porting and preparing a final
> version.
>
> Thanks.
Here it is.
--
Luca Saiu
* My personal web site: http://ageinghacker.net
* GNU epsilon: http://www.gnu.org/software/epsilon
* Jitter: http://ageinghacker.net/projects/jitter
I support everyone's freedom of mocking any opinion or belief, no
matter how deeply held, with open disrespect and the same unrelented
enthusiasm of a toddler who has just learned the word "poo".
commit b30c5527f25a3619fa30968f71eb1ddcc00db886
Author: Luca Saiu <address@hidden>
Date: Sun Dec 1 14:47:31 2019 +0100
low-level PVM optimization and cleanup
* Update jitter submodule.
* src/pkl-insn.def (PKL_INSN_TUCK): Define new instruction.
* src/pvm.jitter (rot, nrot): Simplify definition using Jitter
macro.
(tuck): New instruction.
(revn): Add specialized arguments. Use unsigned JITTER_ARGU0 to
catch bugs more easily.
(swap-over-to-tuck): New rewrite rule.
(rot-swap-to-quake): Re-introduce deleted rewrite rule, which was
actually correct.
* pkl-asm.pks (remap, write, addo, subo, mulo, divo, modo, atrim)
(cdivo, ais, bconc): Simplify and optimize using new instructions.
* src/pkl-gen.pks (struct_field_mapper): Likewise.
diff --git a/ChangeLog b/ChangeLog
index 47db99c..6490c0c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,18 @@
+2019-12-01 Luca Saiu <address@hidden>
+ * src/pkl-insn.def (PKL_INSN_TUCK): Define new instruction.
+ * src/pvm.jitter (rot, nrot): Simplify definition using Jitter
+ macro.
+ (tuck): New instruction.
+ (revn): Add specialized arguments. Use unsigned JITTER_ARGU0 to
+ catch bugs more easily.
+ (swap-over-to-tuck): New rewrite rule.
+ (rot-swap-to-quake): Re-introduce deleted rewrite rule, which was
+ actually correct.
+ * pkl-asm.pks (remap, write, addo, subo, mulo, divo, modo, atrim)
+ (cdivo, ais, bconc): Simplify and optimize using new instructions.
+ * src/pkl-gen.pks (struct_field_mapper): Likewise.
+
+
2019-12-01 John Darrington <address@hidden>
* src/pk_dump.pk (dump): New parameter cluster_by
Submodule jitter 6f53a42..e095305:
> new stack operation bulge
> stack header: move comment to a better position
> new stack operation whirl
> align snippet code to a conservatively safe power of 2
> new stack operation: slide
> stack macros: accept unsigned depths
> uninspired VM: make debugging instructions non-relocatable
> cosmetic changes
> comment changes
> jitterlisp bug fix
> add windows support
> rename misnamed feature macro
> build system sanity check
> lots of portability improvements
> manual change
> comment fix
> make support for .section .note.GNU-stack conditional even on GNU
> cosmetic change
> stack: tentative optimization for over in the TOS case
> new stack primitive: tuck
diff --git a/src/pkl-asm.pks b/src/pkl-asm.pks
index 60e36ea..ed1b263 100644
--- a/src/pkl-asm.pks
+++ b/src/pkl-asm.pks
@@ -58,8 +58,7 @@
mgetsiz ; WCLS MCLS IOS OFF EBOUND VAL SBOUND
swap ; WCLS MCLS IOS OFF EBOUND SBOUND VAL
mgetm ; WCLS MCLS IOS OFF EBOUND SBOUND VAL MCLS
- swap ; WCLS MCLS IOS OFF EBOUND SBOUND MCLS VAL
- drop ; WCLS MCLS IOS OFF EBOUND SBOUND MCLS
+ nip ; WCLS MCLS IOS OFF EBOUND SBOUND MCLS
call ; WCLS MCLS NVAL
swap ; WCLS NVAL MCLS
msetm ; WCLS NVAL
@@ -88,7 +87,7 @@
nrot ; VAL IOS OFF VAL [WCLS]
fromr ; VAL IOS OFF VAL WCLS
call ; VAL null
- push null ; VAL null null
+ dup ; VAL null null
.label:
drop ; VAL (VAL|null)
drop ; VAL
@@ -184,10 +183,9 @@
rot ; OFF1 OFF1M OFF2
ogetm ; OFF1 OFF1M OFF2 OFF2M
rot ; OFF1 OFF2 OFF2M OFF1M
- swap ; OFF1 OFF2 OFF1M OFF2M
add @base_type
- nip2 ; OFF1 OFF2 (OFF1M+OFF2M)
- push #unit ; OFF1 OFF2 (OFF1M+OFF2M) UNIT
+ nip2 ; OFF1 OFF2 (OFF2M+OFF1M)
+ push #unit ; OFF1 OFF2 (OFF2M+OFF1M) UNIT
mko ; OFF1 OFF2 OFFR
.end
@@ -208,8 +206,7 @@
ogetm ; OFF2 OFF1 OFF1M
rot ; OFF1 OFF1M OFF2
ogetm ; OFF1 OFF1M OFF2 OFF2M
- rot ; OFF1 OFF2 OFF2M OFF1M
- swap ; OFF1 OFF2 OFF1M OFF2M
+ quake ; OFF1 OFF2 OFF1M OFF2M
sub @base_type
nip2 ; OFF1 OFF2 (OFF1M+OFF2M)
push #unit ; OFF1 OFF2 (OFF1M+OFF2M) UNIT
@@ -236,8 +233,7 @@
nip2 ; OFF (OFFM*VAL)
swap ; (OFFM*VAL) OFF
ogetu ; (OFFM*VAL) OFF UNIT
- rot ; OFF UNIT (OFFM*VAL)
- swap ; OFF (OFFM*VAL) UNIT
+ quake ; OFF (OFFM*VAL) UNIT
mko ; OFF OFFR
fromr ; OFF OFFR VAL
swap ; OFF VAL OFFR
@@ -259,8 +255,7 @@
ogetm ; OFF2 OFF1 OFF1M
rot ; OFF1 OFF1M OFF2
ogetm ; OFF1 OFF1M OFF2 OFF2M
- rot ; OFF1 OFF2 OFF2M OFF1M
- swap ; OFF1 OFF2 OFF1M OFF2M
+ quake ; OFF1 OFF2 OFF1M OFF2M
div @base_type
nip2 ; OFF1 OFF2 (OFF1M/OFF2M)
.end
@@ -283,8 +278,7 @@
ogetm ; OFF2 OFF1 OFF1M
rot ; OFF1 OFF1M OFF2
ogetm ; OFF1 OFF1M OFF2 OFF2M
- rot ; OFF1 OFF2 OFF2M OFF1M
- swap ; OFF1 OFF2 OFF1M OFF2M
+ quake ; OFF1 OFF2 OFF1M OFF2M
mod @base_type
nip2 ; OFF1 OFF2 (OFF1M%OFF2M)
push #unit ; OFF1 OFF2 (OFF1M%OFF2M) UNIT
@@ -354,8 +348,7 @@
rot ; ... NULL IDX EVAL ARR
drop ; ... NULL IDX EVAL
pushvar $from ; ... NULL IDX EVAL FROM
- rot ; ... NULL EVAL FROM IDX
- swap ; ... NULL EVAL IDX FROM
+ quake ; ... NULL EVAL IDX FROM
sublu
nip2 ; ... NULL EVAL (IDX-FROM)
swap ; ... NULL (IDX-FROM) EVAL
@@ -397,8 +390,7 @@
nip ; TARR BOFFSET ARR BOFF(FROM)
rot ; TARR ARR BOFF(FROM) BOFFSET
dup ; TARR ARR BOFF(FROM) BOFFSET BOFFSET
- rot ; TARR ARR BOFFSET BOFFSET BOFF(FROM)
- swap ; TARR ARR BOFFSET BOFF(FROM) BOFFSET
+ quake ; TARR ARR BOFFSET BOFF(FROM) BOFFSET
sublu
nip2 ; TARR ARR BOFFSET (BOFF(FROM)-BOFFSET)
addlu
@@ -531,8 +523,7 @@
ogetm ; OFF2 OFF1 OFF1M
rot ; OFF1 OFF1M OFF2
ogetm ; OFF1 OFF1M OFF2 OFF2M
- rot ; OFF1 OFF2 OFF2M OFF1M
- swap ; OFF1 OFF2 OFF1M OFF2M
+ quake ; OFF1 OFF2 OFF1M OFF2M
cdiv @type
nip2 ; OFF1 OFF2 (OFF1M/^OFF2M)
.end
@@ -581,17 +572,15 @@
push ulong<64>1
addlu ; SEL ELEM VAL IDX 1UL (IDX+1UL) [ARR NRES]
nip2 ; SEL ELEM VAL NIDX [ARR NRES]
- rot ; SEL VAL NIDX ELEM [ARR NRES]
- drop ; SEL VAL NIDX [ARR NREGS]
- nrot ; NIDX SEL VAL [ARR NREGS]
- swap ; NIDX VAL SEL [ARR NRES]
- rot ; VAL SEL NIDX [ARR NRES]
+ tor ; SEL ELEM VAL [ARR NRES NIDX]
+ nip ; SEL VAL [ARR NRES NIDX]
+ swap ; VAL SEL [ARR NRES NIDX]
+ fromr ; VAL SEL NIDX [ARR NRES]
ba .loop
.foundit:
tor ; SEL ELEM VAL IDX [ARR NRES]
- rot ; SEL VAL IDX ELEM [ARR NRES]
- drop ; SEL VAL IDX [ARR NRES]
- tor ; SEL VAL [ARR NRES IDX]
+ tor ; SEL ELEM VAL [ARR NRES IDX]
+ nip ; SEL VAL [ARR NRES IDX]
swap ; VAL SEL [ARR NRES IDX]
fromr ; VAL SEL IDX [ARR NRES]
dup ; VAL SEL IDX IDX [ARR NRES]
@@ -621,10 +610,9 @@
;;; AST node with the type of the result.
.macro bconc #op2_type_size @op1_type @op2_type @res_type
- dup ; OP1 OP2 OP2
- rot ; OP2 OP2 OP1
- dup ; OP2 OP2 OP1 OP1
- rot ; OP2 OP1 OP1 OP2
+ tuck ; OP2 OP1 OP2
+ over ; OP2 OP1 OP2 OP1
+ swap ; OP2 OP1 OP1 OP2
;; Convert the second operand to the result type.
nton @op2_type, @res_type ; ... OP1 OP2 OP2C
nip ; ... OP1 OP2C
diff --git a/src/pkl-gen.pks b/src/pkl-gen.pks
index 68cfbe5..73203bc 100644
--- a/src/pkl-gen.pks
+++ b/src/pkl-gen.pks
@@ -545,11 +545,9 @@
.e check_struct_field_constraint
;; Calculate the offset marking the end of the field, which is
;; the field's offset plus it's size.
- rot ; STR VAL BOFF
- swap ; STR BOFF VAL
+ quake ; STR BOFF VAL
siz ; STR BOFF VAL SIZ
- rot ; STR VAL SIZ BOFF
- swap ; STR VAL BOFF SIZ
+ quake ; STR VAL BOFF SIZ
addlu
nip ; STR VAL BOFF (BOFF+SIZ)
tor ; STR VAL BOFF
diff --git a/src/pkl-insn.def b/src/pkl-insn.def
index 28a93e3..f33a8d4 100644
--- a/src/pkl-insn.def
+++ b/src/pkl-insn.def
@@ -55,6 +55,7 @@ PKL_DEF_INSN (PKL_INSN_DUP, "", "dup")
PKL_DEF_INSN (PKL_INSN_OVER, "", "over")
PKL_DEF_INSN (PKL_INSN_ROT, "", "rot")
PKL_DEF_INSN (PKL_INSN_NROT, "", "nrot")
+PKL_DEF_INSN (PKL_INSN_TUCK, "", "tuck")
PKL_DEF_INSN (PKL_INSN_SAVER, "r", "saver")
PKL_DEF_INSN (PKL_INSN_RESTORER, "r", "restorer")
PKL_DEF_INSN (PKL_INSN_TOR, "", "tor")
diff --git a/src/pvm.jitter b/src/pvm.jitter
index 94b6118..d05b240 100644
--- a/src/pvm.jitter
+++ b/src/pvm.jitter
@@ -950,31 +950,19 @@ end
instruction rot () # ( A B C -- B C A )
code
- pvm_val a, b, c;
-
- c = JITTER_TOP_STACK ();
- JITTER_DROP_STACK ();
- b = JITTER_TOP_STACK ();
- a = JITTER_UNDER_TOP_STACK ();
-
- JITTER_UNDER_TOP_STACK () = b;
- JITTER_TOP_STACK () = c;
- JITTER_PUSH_STACK (a);
+ JITTER_ROT_STACK ();
end
end
instruction nrot () # ( A B C -- C A B )
code
- pvm_val a, b, c;
-
- c = JITTER_TOP_STACK ();
- JITTER_DROP_STACK ();
- b = JITTER_TOP_STACK ();
- a = JITTER_UNDER_TOP_STACK ();
+ JITTER_MROT_STACK ();
+ end
+end
- JITTER_UNDER_TOP_STACK () = c;
- JITTER_TOP_STACK () = a;
- JITTER_PUSH_STACK (b);
+instruction tuck () # ( A B -- B A B )
+ code
+ JITTER_TUCK_STACK ();
end
end
@@ -1033,9 +1021,9 @@ instruction sel () # ( VAL -- VAL ULONG )
end
end
-instruction revn (?n popf_printer) # ( VAL{N} -- VAL{N} )
+instruction revn (?n 3 4 popf_printer) # ( VAL{N} -- VAL{N} )
code
- JITTER_REVERSE_STACK (JITTER_ARGN0);
+ JITTER_REVERSE_STACK (JITTER_ARGU0);
end
end
@@ -2768,3 +2756,14 @@ into
drop4
end
+rule swap-over-to-tuck rewrite
+ swap; over
+into
+ tuck
+end
+
+rule rot-swap-to-quake rewrite
+ rot; swap
+into
+ quake
+end
signature.asc
Description: PGP signature
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization], Jose E. Marchesi, 2019/12/01
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization], Luca Saiu, 2019/12/01
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization], Jose E. Marchesi, 2019/12/01
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization],
Luca Saiu <=
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization], Jose E. Marchesi, 2019/12/01
- Re: [Patch] PVM low-level cleanup and optimization, good version [Was: Re: [Patch, first version] PVM low-level cleanup and optimization], Luca Saiu, 2019/12/01