commit a7ada8d0d20a88c00a35a489c30c309c20a0b815
parent 321d9c2f70722de7735bff7c6ad54f90d5b5a2da
Author: Ryan Sepassi <rsepassi@gmail.com>
Date: Mon, 27 Apr 2026 17:14:56 -0700
scheme1: quote raw-byte padding for riscv64 M0
Bare `00` tokens in :name_*/:str_*/:msg_* padding tripped riscv64 M0's
Eval_Immediates: a token whose second char is '0' falls through to
express_number, which only accepts %/!/@/~ prefixes and Fails on '0'.
Switch every padding run to '00...' quoted-literal form (already the
project-wide convention per the M0 hex-byte memo).
Diffstat:
| M | scheme1/scheme1.P1pp | | | 251 | ++++++++++++++++++++++++++++++++++++++++--------------------------------------- |
1 file changed, 126 insertions(+), 125 deletions(-)
diff --git a/scheme1/scheme1.P1pp b/scheme1/scheme1.P1pp
@@ -6258,132 +6258,133 @@
})
# Surface names. Length is hard-coded at the call site; no NUL needed
-# because intern takes (ptr, len). Aligned padding via "\0" bytes is
-# fine -- M0 emits ASCII verbatim.
-:name_quote "quote" 00 00
-:name_if "if" 00 00 00 00 00
-:name_lambda "lambda" 00
-:name_define "define" 00
-:name_begin "begin" 00 00
-:name_cond "cond" 00 00 00
-:name_else "else" 00 00 00
-:name_arrow "=>" 00 00 00 00 00
-:name_let "let" 00 00 00 00
-:name_letstar "let*" 00 00 00
-:name_letrec "letrec" 00
+# because intern takes (ptr, len). Padding bytes are written as
+# '00...' quoted-literal form (not bare `00 00`) -- the riscv64 stage0
+# M0 misparses bare hex tokens through express_number/Fail.
+:name_quote "quote" '0000'
+:name_if "if" '0000000000'
+:name_lambda "lambda" '00'
+:name_define "define" '00'
+:name_begin "begin" '0000'
+:name_cond "cond" '000000'
+:name_else "else" '000000'
+:name_arrow "=>" '0000000000'
+:name_let "let" '00000000'
+:name_letstar "let*" '000000'
+:name_letrec "letrec" '00'
:name_letrecstar "letrec*"
-:name_let_values "let-values" 00 00 00 00 00
-:name_letstar_values "let*-values" 00 00 00 00
-:name_define_values "define-values" 00 00
-:name_and "and" 00 00 00 00
-:name_or "or" 00 00 00 00 00
-:name_when "when" 00 00 00
-:name_unless "unless" 00
-:name_case "case" 00 00 00
-:name_setbang "set!" 00 00 00
-:name_define_record_type "define-record-type" 00 00 00 00 00
-:name_pmatch "pmatch" 00
-:name_do "do" 00 00 00 00 00
-:name_case_lambda "case-lambda" 00 00 00 00
+:name_let_values "let-values" '0000000000'
+:name_letstar_values "let*-values" '00000000'
+:name_define_values "define-values" '0000'
+:name_and "and" '00000000'
+:name_or "or" '0000000000'
+:name_when "when" '000000'
+:name_unless "unless" '00'
+:name_case "case" '000000'
+:name_setbang "set!" '000000'
+:name_define_record_type "define-record-type" '0000000000'
+:name_pmatch "pmatch" '00'
+:name_do "do" '0000000000'
+:name_case_lambda "case-lambda" '00000000'
:name_unquote "unquote"
-:name_quasiquote "quasiquote" 00 00 00 00 00
-:name_unquote_splicing "unquote-splicing" 00 00 00 00 00 00 00
-:name_guard "guard" 00 00
-:name_underscore "_" 00 00 00 00 00 00
-:name_dollar "$" 00 00 00 00 00 00
+:name_quasiquote "quasiquote" '0000000000'
+:name_unquote_splicing "unquote-splicing" '00000000000000'
+:name_guard "guard" '0000'
+:name_underscore "_" '000000000000'
+:name_dollar "$" '000000000000'
# Primitive surface names.
-:name_sys_exit "sys-exit" 00 00 00 00 00 00 00
-:name_cons "cons" 00 00 00
-:name_car "car" 00 00 00 00
-:name_cdr "cdr" 00 00 00 00
-:name_nullq "null?" 00 00
-:name_pairq "pair?" 00 00
+:name_sys_exit "sys-exit" '00000000000000'
+:name_cons "cons" '000000'
+:name_car "car" '00000000'
+:name_cdr "cdr" '00000000'
+:name_nullq "null?" '0000'
+:name_pairq "pair?" '0000'
:name_stringq "string?"
-:name_set_car "set-car!" 00 00 00 00 00 00 00
-:name_set_cdr "set-cdr!" 00 00 00 00 00 00 00
-:name_length "length" 00
-:name_list_ref "list-ref" 00 00 00 00 00 00 00
-:name_str_to_sym "string->symbol" 00
-:name_sym_to_str "symbol->string" 00
-:name_num_to_str "number->string" 00
-:name_str_to_num "string->number" 00
-:name_bv_append "bytevector-append" 00 00 00 00 00 00
-:name_booleanq "boolean?" 00 00 00 00 00 00 00
-:name_integerq "integer?" 00 00 00 00 00 00 00
+:name_set_car "set-car!" '00000000000000'
+:name_set_cdr "set-cdr!" '00000000000000'
+:name_length "length" '00'
+:name_list_ref "list-ref" '00000000000000'
+:name_str_to_sym "string->symbol" '00'
+:name_sym_to_str "symbol->string" '00'
+:name_num_to_str "number->string" '00'
+:name_str_to_num "string->number" '00'
+:name_bv_append "bytevector-append" '000000000000'
+:name_booleanq "boolean?" '00000000000000'
+:name_integerq "integer?" '00000000000000'
:name_symbolq "symbol?"
-:name_procedureq "procedure?" 00 00 00 00 00
-:name_zeroq "zero?" 00 00
-:name_not "not" 00 00 00 00
-:name_eqq "eq?" 00 00 00 00
-:name_equal "equal?" 00
-:name_plus "+" 00 00 00 00 00 00
-:name_minus "-" 00 00 00 00 00 00
-:name_mult "*" 00 00 00 00 00 00
-:name_eq "=" 00 00 00 00 00 00
-:name_lt "<" 00 00 00 00 00 00
-:name_gt ">" 00 00 00 00 00 00
-:name_quotient "quotient" 00 00 00 00 00 00 00
-:name_remainder "remainder" 00 00 00 00 00 00
+:name_procedureq "procedure?" '0000000000'
+:name_zeroq "zero?" '0000'
+:name_not "not" '00000000'
+:name_eqq "eq?" '00000000'
+:name_equal "equal?" '00'
+:name_plus "+" '000000000000'
+:name_minus "-" '000000000000'
+:name_mult "*" '000000000000'
+:name_eq "=" '000000000000'
+:name_lt "<" '000000000000'
+:name_gt ">" '000000000000'
+:name_quotient "quotient" '00000000000000'
+:name_remainder "remainder" '000000000000'
:name_bit_and "bit-and"
-:name_bit_or "bit-or" 00
+:name_bit_or "bit-or" '00'
:name_bit_xor "bit-xor"
:name_bit_not "bit-not"
-:name_arith_shift "arithmetic-shift" 00 00 00 00 00 00 00
-:name_apply "apply" 00 00
+:name_arith_shift "arithmetic-shift" '00000000000000'
+:name_apply "apply" '0000'
:name_make_bv "make-bytevector"
-:name_bv_length "bytevector-length" 00 00 00 00 00 00
-:name_string_length "string-length" 00 00
-:name_bv_u8_ref "bytevector-u8-ref" 00 00 00 00 00 00
-:name_bv_u8_set "bytevector-u8-set!" 00 00 00 00 00
+:name_bv_length "bytevector-length" '000000000000'
+:name_string_length "string-length" '0000'
+:name_bv_u8_ref "bytevector-u8-ref" '000000000000'
+:name_bv_u8_set "bytevector-u8-set!" '0000000000'
:name_bv_copy "bytevector-copy"
-:name_bv_copy_b "bytevector-copy!" 00 00 00 00 00 00 00
-:name_bv_eq "bytevector=?" 00 00 00
-
-:name_sys_read "sys-read" 00 00 00 00 00 00 00
-:name_sys_write "sys-write" 00 00 00 00 00 00
-:name_sys_close "sys-close" 00 00 00 00 00 00
-:name_sys_openat "sys-openat" 00 00 00 00 00
-:name_sys_clone "sys-clone" 00 00 00 00 00 00
-:name_sys_execve "sys-execve" 00 00 00 00 00
-:name_sys_waitid "sys-waitid" 00 00 00 00 00
-:name_sys_argv "sys-argv" 00 00 00 00 00 00 00
-:name_eof_object "eof-object" 00 00 00 00 00
-:name_eof_objectq "eof-object?" 00 00 00 00
-
-:name_values "values" 00
-:name_call_with_values "call-with-values" 00 00 00 00 00 00 00
+:name_bv_copy_b "bytevector-copy!" '00000000000000'
+:name_bv_eq "bytevector=?" '000000'
+
+:name_sys_read "sys-read" '00000000000000'
+:name_sys_write "sys-write" '000000000000'
+:name_sys_close "sys-close" '000000000000'
+:name_sys_openat "sys-openat" '0000000000'
+:name_sys_clone "sys-clone" '000000000000'
+:name_sys_execve "sys-execve" '0000000000'
+:name_sys_waitid "sys-waitid" '0000000000'
+:name_sys_argv "sys-argv" '00000000000000'
+:name_eof_object "eof-object" '0000000000'
+:name_eof_objectq "eof-object?" '00000000'
+
+:name_values "values" '00'
+:name_call_with_values "call-with-values" '00000000000000'
:name_display "display"
-:name_write "write" 00 00
-:name_error "error" 00 00
-:name_format "format" 00
+:name_write "write" '0000'
+:name_error "error" '0000'
+:name_format "format" '00'
;; The last three names are padded individually to 16 bytes (multiple
;; of 8) so subsequent 8-aligned data slots (prim_table $() rows, the
;; bss pointer slots) stay aligned. M0 appends a NUL to every "..."
;; string, so the listed length below counts the trailing NUL plus the
;; explicit 00 bytes.
-;; "heap-mark" + NUL = 10 bytes; pad with 6 00 to reach 16.
-;; "heap-rewind!"+ NUL = 13 bytes; pad with 3 00 to reach 16.
-;; "heap-usage" + NUL = 11 bytes; pad with 5 00 to reach 16.
-:name_heap_mark "heap-mark" 00 00 00 00 00 00
-:name_heap_rewind_bang "heap-rewind!" 00 00 00
-:name_heap_usage "heap-usage" 00 00 00 00 00
+;; "heap-mark" + NUL = 10 bytes; pad with 6 NUL to reach 16.
+;; "heap-rewind!"+ NUL = 13 bytes; pad with 3 NUL to reach 16.
+;; "heap-usage" + NUL = 11 bytes; pad with 5 NUL to reach 16.
+:name_heap_mark "heap-mark" '000000000000'
+:name_heap_rewind_bang "heap-rewind!" '000000'
+:name_heap_usage "heap-usage" '0000000000'
# Writer string constants. Lengths are hard-coded at the str_putn call
# sites (write_to_bv branches). No NUL needed in the source bytes --
# str_putn takes (ptr, n).
-:str_false "#f" 00 00 00 00 00 00
-:str_true "#t" 00 00 00 00 00 00
-:str_nil "()" 00 00 00 00 00 00
-:str_unspec "#!unspec" 00 00 00 00 00 00 00 00
-:str_unbound "#!unbound" 00 00 00 00 00 00 00
-:str_eof "#!eof" 00 00 00
-:str_closure "#<closure>" 00 00 00 00 00 00
-:str_prim "#<prim>" 00
-:str_td "#<rec-type>" 00 00 00 00 00
-:str_rec "#<record>" 00 00 00 00 00 00 00
-:str_unknown "#<unknown>" 00 00 00 00 00 00
+:str_false "#f" '000000000000'
+:str_true "#t" '000000000000'
+:str_nil "()" '000000000000'
+:str_unspec "#!unspec" '0000000000000000'
+:str_unbound "#!unbound" '00000000000000'
+:str_eof "#!eof" '000000'
+:str_closure "#<closure>" '000000000000'
+:str_prim "#<prim>" '00'
+:str_td "#<rec-type>" '0000000000'
+:str_rec "#<record>" '00000000000000'
+:str_unknown "#<unknown>" '000000000000'
:str_error_prefix "scheme1: error: "
# Primitive registration table. Each entry: 8-byte name_ptr (4-byte label
@@ -6456,27 +6457,27 @@
&name_call_with_values %(0) $(16) &prim_call_with_values_entry %(0)
:prim_table_end
-:msg_usage "scheme1: usage: scheme1 SOURCE.scm" '0a' 00
-:msg_load_fail "scheme1: failed to read source" '0a' 00
-:msg_symtab_full "scheme1: symbol table full" '0a' 00
-:msg_unexp_rparen "scheme1: unexpected ')'" '0a' 00
-:msg_bad_hash "scheme1: bad #-syntax" '0a' 00
-:msg_unexp_eof "scheme1: unexpected EOF in form" '0a' 00
-:msg_unterm_list "scheme1: unterminated list" '0a' 00
-:msg_unbound "scheme1: unbound variable" '0a' 00
-:msg_not_proc "scheme1: not a procedure" '0a' 00
-:msg_heap_full "scheme1: heap exhausted" '0a' 00
-:msg_readbuf_full "scheme1: source buffer overflow" '0a' 00
-:msg_bv_oob "scheme1: bytevector index out of range" '0a' 00
-:msg_unterm_string "scheme1: unterminated string literal" '0a' 00
-:msg_bad_escape "scheme1: bad string escape" '0a' 00
-:msg_bad_char "scheme1: bad #\\ character literal" '0a' 00
-:msg_bad_number "scheme1: bad number literal" '0a' 00
-:msg_bad_ident "scheme1: bad identifier" '0a' 00
-:msg_internal_define "scheme1: internal define is not supported (use letrec)" '0a' 00
-:msg_pmatch_no_match "scheme1: pmatch: no clause matched" '0a' 00
-:msg_bad_unquote_pattern "scheme1: pmatch: malformed ,-pattern" '0a' 00
-:msg_case_lambda_no_match "scheme1: case-lambda: no clause matched arity" '0a' 00
+:msg_usage "scheme1: usage: scheme1 SOURCE.scm" '0a' '00'
+:msg_load_fail "scheme1: failed to read source" '0a' '00'
+:msg_symtab_full "scheme1: symbol table full" '0a' '00'
+:msg_unexp_rparen "scheme1: unexpected ')'" '0a' '00'
+:msg_bad_hash "scheme1: bad #-syntax" '0a' '00'
+:msg_unexp_eof "scheme1: unexpected EOF in form" '0a' '00'
+:msg_unterm_list "scheme1: unterminated list" '0a' '00'
+:msg_unbound "scheme1: unbound variable" '0a' '00'
+:msg_not_proc "scheme1: not a procedure" '0a' '00'
+:msg_heap_full "scheme1: heap exhausted" '0a' '00'
+:msg_readbuf_full "scheme1: source buffer overflow" '0a' '00'
+:msg_bv_oob "scheme1: bytevector index out of range" '0a' '00'
+:msg_unterm_string "scheme1: unterminated string literal" '0a' '00'
+:msg_bad_escape "scheme1: bad string escape" '0a' '00'
+:msg_bad_char "scheme1: bad #\\ character literal" '0a' '00'
+:msg_bad_number "scheme1: bad number literal" '0a' '00'
+:msg_bad_ident "scheme1: bad identifier" '0a' '00'
+:msg_internal_define "scheme1: internal define is not supported (use letrec)" '0a' '00'
+:msg_pmatch_no_match "scheme1: pmatch: no clause matched" '0a' '00'
+:msg_bad_unquote_pattern "scheme1: pmatch: malformed ,-pattern" '0a' '00'
+:msg_case_lambda_no_match "scheme1: case-lambda: no clause matched arity" '0a' '00'
:name_ch_tab "tab"
:name_ch_null "null"
