JWASM question.

[hutch--]

Since at least some of the guys using UASM used to use Japheth's JWASM, I have a question. Making a simple object module is easy enough to do but I have not succeeded in getting the "-bin" option to work in 64 bit. It works fine in 32 bit so the questions is, is it only for 16 and 32 bit ?

I can get the raw binary output I am after with MASM but its a lot more messing around that what JWASM is capable of doing in 32 bit.

[Vortex]

A quick test with the -bin option :

Code: [Select]

option casemap:none
option win64:3
 
.code

StringLen PROC lpszText:QWORD

    mov     r8,1
    xor     rax,rax
    sub     rax,r8
@@:
    add     rax,r8
    movzx   rdx,BYTE PTR [rcx+rax]
    test    rdx,rdx
    jnz     @b
    ret

StringLen ENDP

END

Code: [Select]

uasm64.exe -bin StringLen.asm
UASM v2.38, Aug  4 2017, Masm-compatible assembler.
Portions Copyright (c) 1992-2002 Sybase, Inc. All Rights Reserved.
Source code is available under the Sybase Open Watcom Public License.

StringLen.asm(4) : Error A2096: Model is not declared
StringLen.asm(6) : Error A2082: Must be in segment block
StringLen.asm(8) : Error A2082: Must be in segment block
StringLen.asm(9) : Error A2082: Must be in segment block
StringLen.asm(10) : Error A2082: Must be in segment block
StringLen.asm(11) : Error A2082: Must be in segment block
StringLen.asm(12) : Error A2082: Must be in segment block
StringLen.asm(13) : Error A2082: Must be in segment block
StringLen.asm(14) : Error A2082: Must be in segment block
StringLen.asm(15) : Error A2082: Must be in segment block
StringLen.asm(16) : Error A2082: Must be in segment block
StringLen.asm(18) : Error A2142: Unmatched block nesting: StringLen
StringLen.asm: 20 lines, 1 passes, 15 ms, 0 warnings, 12 errors

[nidud]

deleted

[Vortex]

Hi nidud,

Thanks for your help. It works as expected :

Code: [Select]

\tdm-gcc-32\bin\objdump -M intel -b binary -D -m i386:x86-64 StringLen.BIN

StringLen.BIN:     file format binary

Disassembly of section .data:

00000000 <.data>:
   0:   48 83 ec 08             sub    rsp,0x8
   4:   49 c7 c0 01 00 00 00    mov    r8,0x1
   b:   48 33 c0                xor    rax,rax
   e:   49 2b c0                sub    rax,r8
  11:   49 03 c0                add    rax,r8
  14:   48 0f b6 14 08          movzx  rdx,BYTE PTR [rax+rcx*1]
  19:   48 85 d2                test   rdx,rdx
  1c:   75 f3                   jne    0x11
  1e:   48 83 c4 08             add    rsp,0x8
  22:   c3                      ret

[nidud]

deleted

[hutch--]

Thanks nidud,

.x64
.model  flat, fastcall

.code
option casemap:none
option win64:3

Works just fine.

; ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤

 NOSTACKFRAME

 slen proc

    db 72,139,193,72,131,232,1,72,131,192,1,76,15,182,16,77
    db 133,210,116,39,72,131,192,1,76,15,182,16,77,133,210,116
    db 26,72,131,192,1,76,15,182,16,77,133,210,116,13,72,131
    db 192,1,76,15,182,16,77,133,210,117,204,72,43,193,195

 slen endp

 STACKFRAME

; ¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤¤

[aw27]

option win64:3
asmc
Code: [Select]
   0:   48 89 4c 24 08          mov    QWORD PTR [rsp+0x8],rcx
...
9:   48 83 ec 20             sub    rsp,0x20

Actually UASM does save the register to the parameter  (mov    QWORD PTR [rsp+0x8],rcx) only if the value is used, otherwise not. Also, UASM does not create shadow space (sub    rsp,0x20) if the function is a leaf.

May be you should think about that nidud (if you want want and have time, of course).

No need to thank me for this info, habran.

[habran]

OK, aw27 Thank you ;)

[hutch--]

Japheth's JWASM 2.11s did the job fine, utility is finished and works correctly.

[nidud]

deleted

[aw27]

My (conservative) thinking is then to preserve this predictability and thus limit the extensions to this logic, so the priorities differ and hence also the result.

Very deep thought indeed.  :t