fix some alignment issues

diff --git a/libpcsxcore/new_dynarec/linkage_arm.s b/libpcsxcore/new_dynarec/linkage_arm.s
index 70836df..7b76aec 100644 (file)
--- a/libpcsxcore/new_dynarec/linkage_arm.s
+++ b/libpcsxcore/new_dynarec/linkage_arm.s
@@ -703,11 +703,23 @@ jump_hlecall:
        ldr     r2, [fp, #last_count-dynarec_local]
        str     r0, [fp, #pcaddr-dynarec_local]
        add     r2, r2, r10
        ldr     r2, [fp, #last_count-dynarec_local]
        str     r0, [fp, #pcaddr-dynarec_local]
        add     r2, r2, r10

-       str     r2, [fp, #cycle-dynarec_local] /* PCSX cycle counter */

        adr     lr, pcsx_return
        adr     lr, pcsx_return

+       str     r2, [fp, #cycle-dynarec_local] /* PCSX cycle counter */

        bx      r1
        .size   jump_hlecall, .-jump_hlecall
 
        bx      r1
        .size   jump_hlecall, .-jump_hlecall
 

+       .align  2
+       .global jump_intcall
+       .type   jump_intcall, %function
+jump_intcall:
+       ldr     r2, [fp, #last_count-dynarec_local]
+       str     r0, [fp, #pcaddr-dynarec_local]
+       add     r2, r2, r10
+       adr     lr, pcsx_return
+       str     r2, [fp, #cycle-dynarec_local] /* PCSX cycle counter */
+       b       execI
+       .size   jump_hlecall, .-jump_hlecall
+

 new_dyna_leave:
        .align  2
        .global new_dyna_leave
 new_dyna_leave:
        .align  2
        .global new_dyna_leave


@@ -822,34 +834,34 @@ ari_read_ram_mirror32:
        ari_read_ram_mirror (3<<11), ldr
 
 /* invalidation is already taken care of by the caller */
        ari_read_ram_mirror (3<<11), ldr
 
 /* invalidation is already taken care of by the caller */

-.macro ari_write_ram bic_const var op
+.macro ari_write_ram bic_const var pf

        ldr     r0, [fp, #address-dynarec_local]
        ldr     r0, [fp, #address-dynarec_local]

-       ldr     r1, [fp, #\var-dynarec_local]
+       ldr\pf  r1, [fp, #\var-dynarec_local]

 .if \bic_const
        bic     r0, r0, #\bic_const
 .endif
 .if \bic_const
        bic     r0, r0, #\bic_const
 .endif

-       \op     r1, [r0]
+       str\pf  r1, [r0]

        mov     pc, lr
 .endm
 
 ari_write_ram8:
        mov     pc, lr
 .endm
 
 ari_write_ram8:

-       ari_write_ram 0, byte, strb
+       ari_write_ram 0, byte, b

 
 ari_write_ram16:
 
 ari_write_ram16:

-       ari_write_ram 1, hword, strh
+       ari_write_ram 1, hword, h

 
 ari_write_ram32:
 
 ari_write_ram32:

-       ari_write_ram 3, word, str
+       ari_write_ram 3, word,

 
 

-.macro ari_write_ram_mirror mvn_const var op
+.macro ari_write_ram_mirror mvn_const var pf

        ldr     r0, [fp, #address-dynarec_local]
        mvn     r3, #\mvn_const
        ldr     r0, [fp, #address-dynarec_local]
        mvn     r3, #\mvn_const

-       ldr     r1, [fp, #\var-dynarec_local]
+       ldr\pf  r1, [fp, #\var-dynarec_local]

        and     r0, r3, lsr #11
        ldr     r2, [fp, #invc_ptr-dynarec_local]
        orr     r0, r0, #1<<31
        ldrb    r2, [r2, r0, lsr #12]
        and     r0, r3, lsr #11
        ldr     r2, [fp, #invc_ptr-dynarec_local]
        orr     r0, r0, #1<<31
        ldrb    r2, [r2, r0, lsr #12]

-       \op     r1, [r0]
+       str\pf  r1, [r0]

        tst     r2, r2
        movne   pc, lr
        lsr     r0, r0, #12
        tst     r2, r2
        movne   pc, lr
        lsr     r0, r0, #12


@@ -857,13 +869,13 @@ ari_write_ram32:
 .endm
 
 ari_write_ram_mirror8:
 .endm
 
 ari_write_ram_mirror8:

-       ari_write_ram_mirror 0, byte, strb
+       ari_write_ram_mirror 0, byte, b

 
 ari_write_ram_mirror16:
 
 ari_write_ram_mirror16:

-       ari_write_ram_mirror (1<<11), hword, strh
+       ari_write_ram_mirror (1<<11), hword, h

 
 ari_write_ram_mirror32:
 
 ari_write_ram_mirror32:

-       ari_write_ram_mirror (3<<11), word, str
+       ari_write_ram_mirror (3<<11), word,

 
 
 .macro ari_read_bios_mirror bic_const op
 
 
 .macro ari_read_bios_mirror bic_const op


@@ -963,9 +975,9 @@ ari_read_io32:
 .endif
 .endm
 
 .endif
 .endm
 

-.macro ari_write_io opvl opst var mem_tab tab_shift
+.macro ari_write_io pf var mem_tab tab_shift

        ldr     r0, [fp, #address-dynarec_local]
        ldr     r0, [fp, #address-dynarec_local]

-       \opvl   r1, [fp, #\var-dynarec_local]
+       ldr\pf  r1, [fp, #\var-dynarec_local]

 .if \tab_shift == 0
        bic     r0, r0, #3
 .endif
 .if \tab_shift == 0
        bic     r0, r0, #3
 .endif


@@ -985,7 +997,7 @@ ari_read_io32:
        bxne    r12
 0:
        ldr     r3, [fp, #psxH_ptr-dynarec_local]
        bxne    r12
 0:
        ldr     r3, [fp, #psxH_ptr-dynarec_local]

-       \opst   r1, [r2, r3]
+       str\pf  r1, [r2, r3]

        mov     pc, lr
 1:
 .if \tab_shift == 1 @ write16
        mov     pc, lr
 1:
 .if \tab_shift == 1 @ write16


@@ -1018,9 +1030,9 @@ ari_write_io8:
        mov     pc, lr
 
 ari_write_io16:
        mov     pc, lr
 
 ari_write_io16:

-       ari_write_io ldrh, strh, hword, tab_write16, 1
+       ari_write_io h, hword, tab_write16, 1

 
 ari_write_io32:
 
 ari_write_io32:

-       ari_write_io ldr, str, word, tab_write32, 0
+       ari_write_io , word, tab_write32, 0

 
 @ vim:filetype=armasm
 
 @ vim:filetype=armasm

diff --git a/plugins/dfxvideo/gpu.c b/plugins/dfxvideo/gpu.c
index 200913c..7372ba4 100644 (file)
--- a/plugins/dfxvideo/gpu.c
+++ b/plugins/dfxvideo/gpu.c
@@ -946,14 +946,14 @@ ENDVRAM:
         {
          gpuDataC = primTableCX[command];
          gpuCommand = command;
         {
          gpuDataC = primTableCX[command];
          gpuCommand = command;

-         PUTLE32(&gpuDataM[0], gdata);
+         PUTLE32_(&gpuDataM[0], gdata);

          gpuDataP = 1;
         }
        else continue;
       }
      else
       {
          gpuDataP = 1;
         }
        else continue;
       }
      else
       {

-       PUTLE32(&gpuDataM[gpuDataP], gdata);
+       PUTLE32_(&gpuDataM[gpuDataP], gdata);

        if(gpuDataC>128)
         {
          if((gpuDataC==254 && gpuDataP>=3) ||
        if(gpuDataC>128)
         {
          if((gpuDataC==254 && gpuDataP>=3) ||


@@ -986,7 +986,7 @@ ENDVRAM:
 
 void CALLBACK GPUwriteData(uint32_t gdata)
 {
 
 void CALLBACK GPUwriteData(uint32_t gdata)
 {

- PUTLE32(&gdata, gdata);
+ PUTLE32_(&gdata, gdata);

  GPUwriteDataMem(&gdata,1);
 }
 
  GPUwriteDataMem(&gdata,1);
 }
 

diff --git a/plugins/dfxvideo/gpu.h b/plugins/dfxvideo/gpu.h
index 740add5..b7b171e 100644 (file)
--- a/plugins/dfxvideo/gpu.h
+++ b/plugins/dfxvideo/gpu.h
@@ -104,10 +104,17 @@
 #define GETLEs32(X) ((int16_t)GETLE32((uint16_t *)X))
 
 #define GETLE16(X) LE2HOST16(*(uint16_t *)X)
 #define GETLEs32(X) ((int16_t)GETLE32((uint16_t *)X))
 
 #define GETLE16(X) LE2HOST16(*(uint16_t *)X)

-#define GETLE32(X) LE2HOST32(*(uint32_t *)X)
+#define GETLE32_(X) LE2HOST32(*(uint32_t *)X)

 #define GETLE16D(X) ({uint32_t val = GETLE32(X); (val<<16 | val >> 16);})
 #define PUTLE16(X, Y) do{*((uint16_t *)X)=HOST2LE16((uint16_t)Y);}while(0)
 #define GETLE16D(X) ({uint32_t val = GETLE32(X); (val<<16 | val >> 16);})
 #define PUTLE16(X, Y) do{*((uint16_t *)X)=HOST2LE16((uint16_t)Y);}while(0)

-#define PUTLE32(X, Y) do{*((uint32_t *)X)=HOST2LE16((uint32_t)Y);}while(0)
+#define PUTLE32_(X, Y) do{*((uint32_t *)X)=HOST2LE16((uint32_t)Y);}while(0)
+#ifdef __arm__
+#define GETLE32(X) (*(uint16_t *)X|(((uint16_t *)X)[1]<<16))
+#define PUTLE32(X, Y) do{*((uint16_t *)X)=(uint32_t)Y;((uint16_t *)X)[1]=(uint32_t)(Y)>>16;}while(0)
+#else
+#define GETLE32 GETLE32_
+#define PUTLE32 PUTLE32_
+#endif

 
 /////////////////////////////////////////////////////////////////////////////
 
 
 /////////////////////////////////////////////////////////////////////////////