January 2021 - cmucl-cvs - mailman3.common-lisp.net

[Git][cmucl/cmucl] Pushed new branch issue-89-unused-var
by Raymond Toy 13 Jan '21

13 Jan '21

Raymond Toy pushed new branch issue-89-unused-var at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-89-unused-var You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][master] 2 commits: Fix #98: Use FSTP instruction with appropriate size
by Raymond Toy 12 Jan '21

12 Jan '21

Raymond Toy pushed to branch master at cmucl / cmucl Commits: 755f7a80 by Raymond Toy at 2021-01-11T20:05:32-08:00 Fix #98: Use FSTP instruction with appropriate size Instead of printing `FSTP` and `FSTPD`, print `FSTP DWORD PTR` and `FSTP QWORD PTR` to store a single-float and double-float value to memory. Clang likes this form better than the original, but gcc accepts both. - - - - - 9c0c857d by Raymond Toy at 2021-01-12T04:27:13+00:00 Merge branch 'issue-98-fstpd-formatting' into 'master' Fix #98: Use FSTP instruction with appropriate size Closes #98 See merge request cmucl/cmucl!65 - - - - - 1 changed file: - src/compiler/x86/insts.lisp Changes: ===================================== src/compiler/x86/insts.lisp ===================================== @@ -2636,7 +2636,8 @@ ;;; store single from st(0) and pop ;;; (define-instruction fstp (segment dest) - (:printer floating-point ((op '(#b001 #b011)))) + (:printer floating-point ((op '(#b001 #b011))) + '('fstp :tab 'dword " " 'ptr " " reg/mem)) (:emitter (cond ((fp-reg-tn-p dest) (emit-byte segment #b11011101) @@ -2648,7 +2649,8 @@ ;;; store double from st(0) and pop ;;; (define-instruction fstpd (segment dest) - (:printer floating-point ((op '(#b101 #b011)))) + (:printer floating-point ((op '(#b101 #b011))) + '('fstp :tab 'qword " " 'ptr " " reg/mem)) (:printer floating-point-fp ((op '(#b101 #b011)))) (:emitter (cond ((fp-reg-tn-p dest) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/5e87a39ad45b1dde4d565a… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/5e87a39ad45b1dde4d565a… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl] Pushed new branch issue-98-fstpd-formatting
by Raymond Toy 12 Jan '21

12 Jan '21

Raymond Toy pushed new branch issue-98-fstpd-formatting at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-98-fstpd-formatting You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][master] 2 commits: Fix #95: Fix disassembly for je and movzx/movsx instructions
by Raymond Toy 11 Jan '21

11 Jan '21

Raymond Toy pushed to branch master at cmucl / cmucl Commits: b5bfb153 by Raymond Toy at 2021-01-11T05:37:39+00:00 Fix #95: Fix disassembly for je and movzx/movsx instructions We were printing `jeq` for the branch on equal instruction, but it should be `je`. This is fixed by changing the order of the options in the `defconstant` `conditions`. The first in the list is the preferred value, so we can write `je`, `jeq`, or `jz`, but these will all get printed out as `je` instead of the previous `jeq`. For `movzx` and `movsx`, we need to specify the size of the memory object to match Intel syntax. This is done by changing the size of the `ext-reg-reg/mem` format so that the reg/mem field has type `sized-reg/mem` instead of just `reg/mem`. Finally, we just added a note that the `break` instruction is really the `int3` instruction. I don't know why it's called `break` instead of `int3`. - - - - - 5e87a39a by Raymond Toy at 2021-01-11T05:37:39+00:00 Merge branch 'issue-95-fix-disassem-syntax' into 'master' Fix #95: Fix disassembly for je and movzx/movsx instructions Closes #95 See merge request cmucl/cmucl!64 - - - - - 2 changed files: - src/bootfiles/21d/boot-2020-04-1.lisp - src/compiler/x86/insts.lisp Changes: ===================================== src/bootfiles/21d/boot-2020-04-1.lisp ===================================== @@ -10,3 +10,8 @@ ;; x86: cross-x86-x86 ;; sparc: cross-sparc-sparc +;; This is also used to easily change the order of x86::conditions +;; constant so that we prefer je instead of jeq. Without a +;; cross-compile we'd need to handle the refefintion of the +;; defconstant in a different way. See issue #95. + ===================================== src/compiler/x86/insts.lisp ===================================== @@ -255,12 +255,15 @@ (= (tn-offset thing) 0))) (eval-when (compile load eval) +;; If a line has more than one value, then these are all synonyms, but +;; the first one is the one that is preferred when printing the +;; condition code out. (defconstant conditions '((:o . 0) (:no . 1) (:b . 2) (:nae . 2) (:c . 2) (:nb . 3) (:ae . 3) (:nc . 3) - (:eq . 4) (:e . 4) (:z . 4) + (:e . 4) (:eq . 4) (:z . 4) (:ne . 5) (:nz . 5) (:be . 6) (:na . 6) (:nbe . 7) (:a . 7) @@ -794,7 +797,7 @@ (op :field (byte 7 1)) (width :field (byte 1 0) :type 'width) (reg/mem :fields (list (byte 2 14) (byte 3 8)) - :type 'reg/mem) + :type 'sized-reg/mem) (reg :field (byte 3 11) :type 'reg) ;; optional fields (imm)) @@ -832,7 +835,10 @@ (disassem:define-instruction-format (accum-reg/mem 16 :include 'reg/mem :default-printer '(:name :tab accum ", " reg/mem)) - (reg/mem :type 'reg/mem) ; don't need a size + ;; This format uses the accumulator, so the size is known; therefore + ;; we don't really need to print out the memory size, but let's do + ;; it for consistency. + (reg/mem :type 'sized-reg/mem) (accum :type 'accum)) ;;; Same as reg-reg/mem, but with a prefix of #b00001111 @@ -843,7 +849,7 @@ (op :field (byte 7 9)) (width :field (byte 1 8) :type 'width) (reg/mem :fields (list (byte 2 22) (byte 3 16)) - :type 'reg/mem) + :type 'sized-reg/mem) (reg :field (byte 3 19) :type 'reg) ;; optional fields (imm)) @@ -865,7 +871,7 @@ (prefix :field (byte 8 0) :value #b00001111) (op :field (byte 8 8)) (reg/mem :fields (list (byte 2 22) (byte 3 16)) - :type 'reg/mem) + :type 'sized-reg/mem) (reg :field (byte 3 19) :type 'reg) ;; optional fields (imm)) @@ -1129,7 +1135,8 @@ (error "Bogus args to XCHG: ~S ~S" operand1 operand2))))))) (define-instruction lea (segment dst src) - (:printer reg-reg/mem ((op #b1000110) (width 1))) + ;; Don't need to print out the width for the LEA instruction + (:printer reg-reg/mem ((op #b1000110) (width 1) (reg/mem nil :type 'reg/mem))) (:emitter (assert (dword-reg-p dst)) (emit-byte segment #b10001101) @@ -2112,6 +2119,7 @@ (nt "Function end breakpoint trap")) ))) +;; This is really the int3 instruction. (define-instruction break (segment code) (:declare (type (unsigned-byte 8) code)) (:printer byte-imm ((op #b11001100)) '(:name :tab code) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/5b5082c6cbe682d5805630… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/5b5082c6cbe682d5805630… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-95-fix-disassem-syntax] Make more mem access print out the size
by Raymond Toy 10 Jan '21

10 Jan '21

Raymond Toy pushed to branch issue-95-fix-disassem-syntax at cmucl / cmucl Commits: a99465d8 by Raymond Toy at 2021-01-10T11:06:53-08:00 Make more mem access print out the size But we don't want to do that for the LEA instruction. gcc doesn't require it and objdump doesn't print out a size. - - - - - 1 changed file: - src/compiler/x86/insts.lisp Changes: ===================================== src/compiler/x86/insts.lisp ===================================== @@ -797,7 +797,7 @@ (op :field (byte 7 1)) (width :field (byte 1 0) :type 'width) (reg/mem :fields (list (byte 2 14) (byte 3 8)) - :type 'reg/mem) + :type 'sized-reg/mem) (reg :field (byte 3 11) :type 'reg) ;; optional fields (imm)) @@ -835,7 +835,10 @@ (disassem:define-instruction-format (accum-reg/mem 16 :include 'reg/mem :default-printer '(:name :tab accum ", " reg/mem)) - (reg/mem :type 'reg/mem) ; don't need a size + ;; This format uses the accumulator, so the size is known; therefore + ;; we don't really need to print out the memory size, but let's do + ;; it for consistency. + (reg/mem :type 'sized-reg/mem) (accum :type 'accum)) ;;; Same as reg-reg/mem, but with a prefix of #b00001111 @@ -1132,7 +1135,8 @@ (error "Bogus args to XCHG: ~S ~S" operand1 operand2))))))) (define-instruction lea (segment dst src) - (:printer reg-reg/mem ((op #b1000110) (width 1))) + ;; Don't need to print out the width for the LEA instruction + (:printer reg-reg/mem ((op #b1000110) (width 1) (reg/mem nil :type 'reg/mem))) (:emitter (assert (dword-reg-p dst)) (emit-byte segment #b10001101) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/a99465d8e3c6cb2f245b278… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/a99465d8e3c6cb2f245b278… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl] Pushed new branch issue-95-fix-disassem-syntax
by Raymond Toy 10 Jan '21

10 Jan '21

Raymond Toy pushed new branch issue-95-fix-disassem-syntax at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-95-fix-disassem-syn… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][master] 2 commits: Fix #91: Handle loop destructuring
by Raymond Toy 10 Jan '21

10 Jan '21

Raymond Toy pushed to branch master at cmucl / cmucl Commits: 1144015f by Raymond Toy at 2021-01-10T04:33:03+00:00 Fix #91: Handle loop destructuring The destructuring shortcut in loop doesn't require all the parts be available. If they're not, each item is replaced by NIL. This is fixed by still using `destructuring-bind`, except we mark everything as optional. - - - - - 5b5082c6 by Raymond Toy at 2021-01-10T04:33:04+00:00 Merge branch 'issue-91-loop-destructuring-bind' into 'master' Fix #91: loop destructuring bind Closes #91 See merge request cmucl/cmucl!63 - - - - - 1 changed file: - src/code/loop.lisp Changes: ===================================== src/code/loop.lisp ===================================== @@ -995,7 +995,10 @@ collected result will be returned as the value of the LOOP." (if crocks (let ((*ignores* ())) (declare (special *ignores*)) - `((destructuring-bind ,(subst-gensyms-for-nil (car crocks)) + ;; Destructuring in loop doesn't require that the values be + ;; available. The missing elements are filled with NIL. So, + ;; make everything &optional + `((destructuring-bind (&optional ,@(subst-gensyms-for-nil (car crocks))) ,(cadr crocks) (declare (ignore ,@*ignores*)) ,@(loop-build-destructuring-bindings (cddr crocks) forms)))) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/fdeafbc630fd3599062a68… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/fdeafbc630fd3599062a68… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl] Pushed new branch issue-91-loop-destructuring-bind
by Raymond Toy 07 Jan '21

07 Jan '21

Raymond Toy pushed new branch issue-91-loop-destructuring-bind at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-91-loop-destructuri… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][master] 2 commits: Fix #93: Use our fork of ansi-test
by Raymond Toy 07 Jan '21

07 Jan '21

Raymond Toy pushed to branch master at cmucl / cmucl Commits: f993ea43 by Raymond Toy at 2021-01-06T17:11:03-08:00 Fix #93: Use our fork of ansi-test Instead of using the upstream ansi-test for CI, use our own fork. - - - - - fdeafbc6 by Raymond Toy at 2021-01-07T05:36:53+00:00 Merge branch 'issue-93-update-ansi-test-location' into 'master' Fix #93: Use our fork of ansi-test Closes #93 See merge request cmucl/cmucl!62 - - - - - 1 changed file: - .gitlab-ci.yml Changes: ===================================== .gitlab-ci.yml ===================================== @@ -57,7 +57,7 @@ linux:test: - job: linux:build artifacts: true before_script: - - git clone https://gitlab.common-lisp.net/ansi-test/ansi-test.git + - git clone https://gitlab.common-lisp.net/cmucl/ansi-test.git - (cd ansi-test; git checkout rtoy-cmucl-expected-failures) script: - bin/run-tests.sh -l dist/bin/lisp 2>&1 | tee test.log @@ -128,7 +128,7 @@ osx:test: - job: osx:build artifacts: true before_script: - - git clone https://gitlab.common-lisp.net/ansi-test/ansi-test.git + - git clone https://gitlab.common-lisp.net/cmucl/ansi-test.git - (cd ansi-test; git checkout rtoy-cmucl-expected-failures) script: - bin/run-tests.sh -l dist/bin/lisp 2>&1 | tee test.log View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/18a3be34f2a3cadebc9949… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/18a3be34f2a3cadebc9949… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl] Pushed new branch issue-93-update-ansi-test-location
by Raymond Toy 07 Jan '21

07 Jan '21

Raymond Toy pushed new branch issue-93-update-ansi-test-location at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-93-update-ansi-test… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0