[8u] RFR: Initial backport of AArch32 platform support from OpenJDK 9 to OpenJDK 8u

Triple Yang triple.yang at linaro.org
Sat Jan 16 03:26:23 UTC 2016


Hi,

I make a patch to adjust all the misalignment I found except for some of
that in entry(CodeBuffer *cb) in file assembler_aarch32.cpp, about which I
am not sure.

Best wishes.

On 5 January 2016 at 21:51, Edward Nevill <edward.nevill at gmail.com> wrote:

> Hi,
>
> I used the 'expand' command which expands tabs to multiples of 8.
>
> The default indentation in OpenJDK is indeed 2, but this is separate
> from the tab character which usually tabs to multiples of 8, unless you
> set your editor/terminal to a different setting.
>
> Are you aware of specific instances where the source has been
> misformatted due to tab expansion?
>
> All the best,
> Ed.
>
> On Sun, 2016-01-03 at 22:38 +0800, Triple Yang wrote:
> > Hi, Ed,
> >
> >
> > It seems you replace a tab with 8 spaces in this changeset, wheras 2
> > space are most often used for alignment.
> > Would you please confirm it? Thank you.
> >
> >
>
>
>
>
>
-------------- next part --------------
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/assembler_aarch32.cpp
--- a/src/cpu/aarch32/vm/assembler_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/assembler_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -85,615 +85,615 @@
     __ bind(back);
 
 // ThreeRegSft
-    __ add(r8, r2, r11, ::lsr(10));                    //       add r8, r2, r11, lsr #10
-    __ adds(r1, r3, r7, ::asr(1), Assembler::EQ);      //       addEQs r1, r3, r7, asr #1
-    __ eor(r0, r9, r4, ::lsl(5));                      //       eor r0, r9, r4, lsl #5
-    __ eors(r9, r2, r6, ::rrx(), Assembler::GT);       //       eorGTs r9, r2, r6, rrx
-    __ sub(r0, r12, lr, ::lsr(0), Assembler::GT);      //       subGT r0, r12, lr, lsr #0
-    __ subs(r8, r2, r4, ::ror(6), Assembler::EQ);      //       subEQs r8, r2, r4, ror #6
-    __ rsb(r8, r9, sp, ::lsl(3));                      //       rsb r8, r9, sp, lsl #3
-    __ rsbs(r8, r0, r4, ::ror(16), Assembler::VS);     //       rsbVSs r8, r0, r4, ror #16
-    __ add(r9, r5, r1, ::lsr(15), Assembler::LE);      //       addLE r9, r5, r1, lsr #15
-    __ adds(r1, sp, r6, ::asr(5));                     //       adds r1, sp, r6, asr #5
-    __ adc(r11, sp, r7, ::asr(1), Assembler::GT);      //       adcGT r11, sp, r7, asr #1
-    __ adcs(r0, r8, r9, ::lsr(6));                     //       adcs r0, r8, r9, lsr #6
-    __ sbc(r9, r3, r6, ::ror(5));                      //       sbc r9, r3, r6, ror #5
-    __ sbcs(r1, sp, r5, ::asr(16), Assembler::HI);     //       sbcHIs r1, sp, r5, asr #16
-    __ rsc(r8, r2, r6, ::lsl(9), Assembler::CC);       //       rscCC r8, r2, r6, lsl #9
-    __ rscs(r10, r4, sp, ::ror(14));                   //       rscs r10, r4, sp, ror #14
-    __ orr(r11, sp, r5, ::lsl(15), Assembler::NE);     //       orrNE r11, sp, r5, lsl #15
-    __ orrs(r9, r10, r4, ::ror(14));                   //       orrs r9, r10, r4, ror #14
-    __ bic(r9, sp, r5, ::ror(1));                      //       bic r9, sp, r5, ror #1
-    __ bics(r0, r2, r7, ::asr(10));                    //       bics r0, r2, r7, asr #10
+    __ add(r8, r2, r11, ::lsr(10));                    // add r8, r2, r11, lsr #10
+    __ adds(r1, r3, r7, ::asr(1), Assembler::EQ);      // addEQs r1, r3, r7, asr #1
+    __ eor(r0, r9, r4, ::lsl(5));                      // eor r0, r9, r4, lsl #5
+    __ eors(r9, r2, r6, ::rrx(), Assembler::GT);       // eorGTs r9, r2, r6, rrx
+    __ sub(r0, r12, lr, ::lsr(0), Assembler::GT);      // subGT r0, r12, lr, lsr #0
+    __ subs(r8, r2, r4, ::ror(6), Assembler::EQ);      // subEQs r8, r2, r4, ror #6
+    __ rsb(r8, r9, sp, ::lsl(3));                      // rsb r8, r9, sp, lsl #3
+    __ rsbs(r8, r0, r4, ::ror(16), Assembler::VS);     // rsbVSs r8, r0, r4, ror #16
+    __ add(r9, r5, r1, ::lsr(15), Assembler::LE);      // addLE r9, r5, r1, lsr #15
+    __ adds(r1, sp, r6, ::asr(5));                     // adds r1, sp, r6, asr #5
+    __ adc(r11, sp, r7, ::asr(1), Assembler::GT);      // adcGT r11, sp, r7, asr #1
+    __ adcs(r0, r8, r9, ::lsr(6));                     // adcs r0, r8, r9, lsr #6
+    __ sbc(r9, r3, r6, ::ror(5));                      // sbc r9, r3, r6, ror #5
+    __ sbcs(r1, sp, r5, ::asr(16), Assembler::HI);     // sbcHIs r1, sp, r5, asr #16
+    __ rsc(r8, r2, r6, ::lsl(9), Assembler::CC);       // rscCC r8, r2, r6, lsl #9
+    __ rscs(r10, r4, sp, ::ror(14));                   // rscs r10, r4, sp, ror #14
+    __ orr(r11, sp, r5, ::lsl(15), Assembler::NE);     // orrNE r11, sp, r5, lsl #15
+    __ orrs(r9, r10, r4, ::ror(14));                   // orrs r9, r10, r4, ror #14
+    __ bic(r9, sp, r5, ::ror(1));                      // bic r9, sp, r5, ror #1
+    __ bics(r0, r2, r7, ::asr(10));                    // bics r0, r2, r7, asr #10
 
 // ThreeRegRSR
-    __ add(sp, r6, r7, ::ror(r7));                     //       add sp, r6, r7, ror r7
-    __ adds(r4, r12, r6, ::ror(r7), Assembler::HI);    //       addHIs r4, r12, r6, ror r7
-    __ eor(r5, r6, r7, ::asr(r12), Assembler::LS);     //       eorLS r5, r6, r7, asr r12
-    __ eors(r8, r5, sp, ::lsl(r4), Assembler::AL);     //       eorALs r8, r5, sp, lsl r4
-    __ sub(r2, r12, r5, ::asr(r0));                    //       sub r2, r12, r5, asr r0
-    __ subs(r9, r3, r7, ::lsl(r12), Assembler::HS);    //       subHSs r9, r3, r7, lsl r12
-    __ rsb(r9, r12, r4, ::lsl(r6), Assembler::GT);     //       rsbGT r9, r12, r4, lsl r6
-    __ rsbs(r8, r2, r12, ::lsl(r1));                   //       rsbs r8, r2, r12, lsl r1
-    __ add(r4, r12, sp, ::lsl(sp));                    //       add r4, r12, sp, lsl sp
-    __ adds(r8, r11, r6, ::ror(sp));                   //       adds r8, r11, r6, ror sp
-    __ adc(r0, r2, r5, ::lsl(r4), Assembler::NE);      //       adcNE r0, r2, r5, lsl r4
-    __ adcs(r11, lr, r6, ::asr(r2));                   //       adcs r11, lr, r6, asr r2
-    __ sbc(r8, r10, lr, ::asr(r3), Assembler::HI);     //       sbcHI r8, r10, lr, asr r3
-    __ sbcs(r1, r12, r5, ::lsl(r6));                   //       sbcs r1, r12, r5, lsl r6
-    __ rsc(r4, r5, lr, ::ror(r10), Assembler::VS);     //       rscVS r4, r5, lr, ror r10
-    __ rscs(r1, r12, sp, ::lsl(r8));                   //       rscs r1, r12, sp, lsl r8
-    __ orr(r8, r1, r6, ::ror(r0), Assembler::VS);      //       orrVS r8, r1, r6, ror r0
-    __ orrs(r11, sp, r7, ::ror(r5));                   //       orrs r11, sp, r7, ror r5
-    __ bic(r4, lr, r6, ::lsl(r2), Assembler::AL);      //       bicAL r4, lr, r6, lsl r2
-    __ bics(r10, r11, sp, ::lsl(r3));                  //       bics r10, r11, sp, lsl r3
+    __ add(sp, r6, r7, ::ror(r7));                     // add sp, r6, r7, ror r7
+    __ adds(r4, r12, r6, ::ror(r7), Assembler::HI);    // addHIs r4, r12, r6, ror r7
+    __ eor(r5, r6, r7, ::asr(r12), Assembler::LS);     // eorLS r5, r6, r7, asr r12
+    __ eors(r8, r5, sp, ::lsl(r4), Assembler::AL);     // eorALs r8, r5, sp, lsl r4
+    __ sub(r2, r12, r5, ::asr(r0));                    // sub r2, r12, r5, asr r0
+    __ subs(r9, r3, r7, ::lsl(r12), Assembler::HS);    // subHSs r9, r3, r7, lsl r12
+    __ rsb(r9, r12, r4, ::lsl(r6), Assembler::GT);     // rsbGT r9, r12, r4, lsl r6
+    __ rsbs(r8, r2, r12, ::lsl(r1));                   // rsbs r8, r2, r12, lsl r1
+    __ add(r4, r12, sp, ::lsl(sp));                    // add r4, r12, sp, lsl sp
+    __ adds(r8, r11, r6, ::ror(sp));                   // adds r8, r11, r6, ror sp
+    __ adc(r0, r2, r5, ::lsl(r4), Assembler::NE);      // adcNE r0, r2, r5, lsl r4
+    __ adcs(r11, lr, r6, ::asr(r2));                   // adcs r11, lr, r6, asr r2
+    __ sbc(r8, r10, lr, ::asr(r3), Assembler::HI);     // sbcHI r8, r10, lr, asr r3
+    __ sbcs(r1, r12, r5, ::lsl(r6));                   // sbcs r1, r12, r5, lsl r6
+    __ rsc(r4, r5, lr, ::ror(r10), Assembler::VS);     // rscVS r4, r5, lr, ror r10
+    __ rscs(r1, r12, sp, ::lsl(r8));                   // rscs r1, r12, sp, lsl r8
+    __ orr(r8, r1, r6, ::ror(r0), Assembler::VS);      // orrVS r8, r1, r6, ror r0
+    __ orrs(r11, sp, r7, ::ror(r5));                   // orrs r11, sp, r7, ror r5
+    __ bic(r4, lr, r6, ::lsl(r2), Assembler::AL);      // bicAL r4, lr, r6, lsl r2
+    __ bics(r10, r11, sp, ::lsl(r3));                  // bics r10, r11, sp, lsl r3
 
 // TwoRegImm
-    __ add(r8, sp, (unsigned)268435462U, Assembler::HI); //     addHI r8, sp, #268435462
-    __ adds(sp, lr, (unsigned)162529280U);             //       adds sp, lr, #162529280
-    __ eor(lr, r6, (unsigned)8192000U);                //       eor lr, r6, #8192000
-    __ eors(r2, r3, (unsigned)292U);                   //       eors r2, r3, #292
-    __ sub(r4, sp, (unsigned)227540992U);              //       sub r4, sp, #227540992
-    __ subs(r1, lr, (unsigned)33554432U, Assembler::LT); //     subLTs r1, lr, #33554432
-    __ rsb(r0, r5, (unsigned)2483027968U);             //       rsb r0, r5, #2483027968
-    __ rsbs(r8, r4, (unsigned)3080192U, Assembler::LO); //      rsbLOs r8, r4, #3080192
-    __ add(r9, r4, (unsigned)2147483648U, Assembler::LT); //    addLT r9, r4, #2147483648
-    __ adds(r8, r4, (unsigned)32768U, Assembler::AL);  //       addALs r8, r4, #32768
-    __ adc(r10, lr, (unsigned)10752U, Assembler::CS);  //       adcCS r10, lr, #10752
-    __ adcs(r10, r6, (unsigned)774144U);               //       adcs r10, r6, #774144
-    __ sbc(r2, r12, (unsigned)637534208U);             //       sbc r2, r12, #637534208
-    __ sbcs(r8, r10, (unsigned)692060160U);            //       sbcs r8, r10, #692060160
-    __ rsc(sp, r6, (unsigned)7405568U);                //       rsc sp, r6, #7405568
-    __ rscs(r10, r11, (unsigned)244318208U, Assembler::NE); //  rscNEs r10, r11, #244318208
-    __ orr(r3, r7, (unsigned)66846720U, Assembler::VS); //      orrVS r3, r7, #66846720
-    __ orrs(r2, r5, (unsigned)1327104U, Assembler::EQ); //      orrEQs r2, r5, #1327104
-    __ bic(r8, r1, (unsigned)3744U, Assembler::VS);    //       bicVS r8, r1, #3744
-    __ bics(r0, r2, (unsigned)2684354560U, Assembler::LO); //   bicLOs r0, r2, #2684354560
+    __ add(r8, sp, (unsigned)268435462U, Assembler::HI); // addHI r8, sp, #268435462
+    __ adds(sp, lr, (unsigned)162529280U);             // adds sp, lr, #162529280
+    __ eor(lr, r6, (unsigned)8192000U);                // eor lr, r6, #8192000
+    __ eors(r2, r3, (unsigned)292U);                   // eors r2, r3, #292
+    __ sub(r4, sp, (unsigned)227540992U);              // sub r4, sp, #227540992
+    __ subs(r1, lr, (unsigned)33554432U, Assembler::LT); // subLTs r1, lr, #33554432
+    __ rsb(r0, r5, (unsigned)2483027968U);             // rsb r0, r5, #2483027968
+    __ rsbs(r8, r4, (unsigned)3080192U, Assembler::LO); // rsbLOs r8, r4, #3080192
+    __ add(r9, r4, (unsigned)2147483648U, Assembler::LT); // addLT r9, r4, #2147483648
+    __ adds(r8, r4, (unsigned)32768U, Assembler::AL);  // addALs r8, r4, #32768
+    __ adc(r10, lr, (unsigned)10752U, Assembler::CS);  // adcCS r10, lr, #10752
+    __ adcs(r10, r6, (unsigned)774144U);               // adcs r10, r6, #774144
+    __ sbc(r2, r12, (unsigned)637534208U);             // sbc r2, r12, #637534208
+    __ sbcs(r8, r10, (unsigned)692060160U);            // sbcs r8, r10, #692060160
+    __ rsc(sp, r6, (unsigned)7405568U);                // rsc sp, r6, #7405568
+    __ rscs(r10, r11, (unsigned)244318208U, Assembler::NE); // rscNEs r10, r11, #244318208
+    __ orr(r3, r7, (unsigned)66846720U, Assembler::VS); // orrVS r3, r7, #66846720
+    __ orrs(r2, r5, (unsigned)1327104U, Assembler::EQ); // orrEQs r2, r5, #1327104
+    __ bic(r8, r1, (unsigned)3744U, Assembler::VS);    // bicVS r8, r1, #3744
+    __ bics(r0, r2, (unsigned)2684354560U, Assembler::LO); // bicLOs r0, r2, #2684354560
 
 // TwoRegSft
-    __ tst(r8, sp, ::lsl(5));                          //       tst r8, sp, lsl #5
-    __ teq(r6, r7, ::lsr(3));                          //       teq r6, r7, lsr #3
-    __ cmp(r12, r4, ::ror(2));                         //       cmp r12, r4, ror #2
-    __ cmn(r5, r7, ::lsl(16), Assembler::LT);          //       cmnLT r5, r7, lsl #16
+    __ tst(r8, sp, ::lsl(5));                          // tst r8, sp, lsl #5
+    __ teq(r6, r7, ::lsr(3));                          // teq r6, r7, lsr #3
+    __ cmp(r12, r4, ::ror(2));                         // cmp r12, r4, ror #2
+    __ cmn(r5, r7, ::lsl(16), Assembler::LT);          // cmnLT r5, r7, lsl #16
 
 // TwoRegRSR
-    __ tst(r2, lr, ::lsr(r7));                         //       tst r2, lr, lsr r7
-    __ teq(r0, r2, ::ror(r5), Assembler::CC);          //       teqCC r0, r2, ror r5
-    __ cmp(lr, r7, ::lsr(r11), Assembler::LS);         //       cmpLS lr, r7, lsr r11
-    __ cmn(r10, r7, ::lsl(r11), Assembler::VS);        //       cmnVS r10, r7, lsl r11
+    __ tst(r2, lr, ::lsr(r7));                         // tst r2, lr, lsr r7
+    __ teq(r0, r2, ::ror(r5), Assembler::CC);          // teqCC r0, r2, ror r5
+    __ cmp(lr, r7, ::lsr(r11), Assembler::LS);         // cmpLS lr, r7, lsr r11
+    __ cmn(r10, r7, ::lsl(r11), Assembler::VS);        // cmnVS r10, r7, lsl r11
 
 // OneRegImm
-    __ tst(r2, (unsigned)557842432U);                  //       tst r2, #557842432
-    __ teq(lr, (unsigned)7077888U, Assembler::MI);     //       teqMI lr, #7077888
-    __ cmp(r5, (unsigned)939524096U);                  //       cmp r5, #939524096
-    __ cmn(r7, (unsigned)2147483650U, Assembler::LO);  //       cmnLO r7, #2147483650
+    __ tst(r2, (unsigned)557842432U);                  // tst r2, #557842432
+    __ teq(lr, (unsigned)7077888U, Assembler::MI);     // teqMI lr, #7077888
+    __ cmp(r5, (unsigned)939524096U);                  // cmp r5, #939524096
+    __ cmn(r7, (unsigned)2147483650U, Assembler::LO);  // cmnLO r7, #2147483650
 
 // Shift op
-    __ lsl(r0, r4, (unsigned)23U);                     //       lsl r0, r4, #23
-    __ lsls(r1, r4, (unsigned)9U);                     //       lsls r1, r4, #9
-    __ lsr(r0, r10, (unsigned)3U);                     //       lsr r0, r10, #3
-    __ lsrs(r0, r10, (unsigned)20U);                   //       lsrs r0, r10, #20
-    __ asr(r1, r9, (unsigned)11U);                     //       asr r1, r9, #11
-    __ asrs(r2, r11, (unsigned)10U, Assembler::VS);    //       asrVSs r2, r11, #10
+    __ lsl(r0, r4, (unsigned)23U);                     // lsl r0, r4, #23
+    __ lsls(r1, r4, (unsigned)9U);                     // lsls r1, r4, #9
+    __ lsr(r0, r10, (unsigned)3U);                     // lsr r0, r10, #3
+    __ lsrs(r0, r10, (unsigned)20U);                   // lsrs r0, r10, #20
+    __ asr(r1, r9, (unsigned)11U);                     // asr r1, r9, #11
+    __ asrs(r2, r11, (unsigned)10U, Assembler::VS);    // asrVSs r2, r11, #10
 
 // shift op
-    __ ror(r8, r2, (unsigned)31U, Assembler::CC);      //       rorCC r8, r2, #31
-    __ rors(r9, r12, (unsigned)8U);                    //       rors r9, r12, #8
+    __ ror(r8, r2, (unsigned)31U, Assembler::CC);      // rorCC r8, r2, #31
+    __ rors(r9, r12, (unsigned)8U);                    // rors r9, r12, #8
 
 // ThreeRegNon
-    __ ror(r8, lr, r7);                                //       ror r8, lr, r7
-    __ rors(r12, r3, r4);                              //       rors r12, r3, r4
-    __ lsl(r12, sp, lr, Assembler::GT);                //       lslGT r12, sp, lr
-    __ lsls(r12, sp, r6, Assembler::AL);               //       lslALs r12, sp, r6
-    __ lsr(r0, r1, r9, Assembler::GT);                 //       lsrGT r0, r1, r9
-    __ lsrs(r11, r3, r12, Assembler::GT);              //       lsrGTs r11, r3, r12
-    __ asr(r2, r12, r6, Assembler::LE);                //       asrLE r2, r12, r6
-    __ asrs(r1, r10, r6, Assembler::LT);               //       asrLTs r1, r10, r6
+    __ ror(r8, lr, r7);                                // ror r8, lr, r7
+    __ rors(r12, r3, r4);                              // rors r12, r3, r4
+    __ lsl(r12, sp, lr, Assembler::GT);                // lslGT r12, sp, lr
+    __ lsls(r12, sp, r6, Assembler::AL);               // lslALs r12, sp, r6
+    __ lsr(r0, r1, r9, Assembler::GT);                 // lsrGT r0, r1, r9
+    __ lsrs(r11, r3, r12, Assembler::GT);              // lsrGTs r11, r3, r12
+    __ asr(r2, r12, r6, Assembler::LE);                // asrLE r2, r12, r6
+    __ asrs(r1, r10, r6, Assembler::LT);               // asrLTs r1, r10, r6
 
 // TwoRegNon
-    __ mov(r10, r3);                                   //       mov r10, r3
-    __ movs(r0, r9);                                   //       movs r0, r9
+    __ mov(r10, r3);                                   // mov r10, r3
+    __ movs(r0, r9);                                   // movs r0, r9
 
 // OneRegImm
-    __ mov_i(r3, (unsigned)656U, Assembler::VC);         //     movVC r3, #656
-    __ movs_i(r4, (unsigned)2064384U);                   //     movs r4, #2064384
+    __ mov_i(r3, (unsigned)656U, Assembler::VC);         // movVC r3, #656
+    __ movs_i(r4, (unsigned)2064384U);                   // movs r4, #2064384
 
 // TwoRegSft
-    __ mov(r12, r6, ::lsr(3));                         //       mov r12, r6, lsr #3
-    __ movs(r5, sp, ::asr(10), Assembler::VC);         //       movVCs r5, sp, asr #10
+    __ mov(r12, r6, ::lsr(3));                         // mov r12, r6, lsr #3
+    __ movs(r5, sp, ::asr(10), Assembler::VC);         // movVCs r5, sp, asr #10
 
 // TwoRegRSR
-    __ mov(r1, lr, ::ror(r3));                         //       mov r1, lr, ror r3
-    __ movs(r8, r12, ::ror(r9), Assembler::EQ);        //       movEQs r8, r12, ror r9
+    __ mov(r1, lr, ::ror(r3));                         // mov r1, lr, ror r3
+    __ movs(r8, r12, ::ror(r9), Assembler::EQ);        // movEQs r8, r12, ror r9
 
 // OneRegImm16
-    __ movw_i(r11, (unsigned)53041U, Assembler::LO);     //     movwLO r11, #53041
-    __ movt_i(r9, (unsigned)11255U, Assembler::LO);      //     movtLO r9, #11255
+    __ movw_i(r11, (unsigned)53041U, Assembler::LO);     // movwLO r11, #53041
+    __ movt_i(r9, (unsigned)11255U, Assembler::LO);      // movtLO r9, #11255
 
 // ThreeRegNon
-    __ mul(r1, sp, r5, Assembler::LE);                 //       mulLE r1, sp, r5
-    __ muls(r0, r10, r11);                             //       muls r0, r10, r11
+    __ mul(r1, sp, r5, Assembler::LE);                 // mulLE r1, sp, r5
+    __ muls(r0, r10, r11);                             // muls r0, r10, r11
 
 // FourRegNon
-    __ mla(r0, r3, r12, r7);                           //       mla r0, r3, r12, r7
-    __ mlas(r8, r11, r3, r6, Assembler::EQ);           //       mlaEQs r8, r11, r3, r6
-    __ umull(lr, r4, r5, r6);                          //       umull lr, r4, r5, r6
-    __ umulls(r0, r4, r6, r7);                         //       umulls r0, r4, r6, r7
-    __ umlal(r8, r0, r11, lr);                         //       umlal r8, r0, r11, lr
-    __ umlals(r11, r4, lr, r7);                        //       umlals r11, r4, lr, r7
-    __ smull(r1, r5, r6, r7, Assembler::HS);           //       smullHS r1, r5, r6, r7
-    __ smulls(r0, r11, r12, r5, Assembler::MI);        //       smullMIs r0, r11, r12, r5
+    __ mla(r0, r3, r12, r7);                           // mla r0, r3, r12, r7
+    __ mlas(r8, r11, r3, r6, Assembler::EQ);           // mlaEQs r8, r11, r3, r6
+    __ umull(lr, r4, r5, r6);                          // umull lr, r4, r5, r6
+    __ umulls(r0, r4, r6, r7);                         // umulls r0, r4, r6, r7
+    __ umlal(r8, r0, r11, lr);                         // umlal r8, r0, r11, lr
+    __ umlals(r11, r4, lr, r7);                        // umlals r11, r4, lr, r7
+    __ smull(r1, r5, r6, r7, Assembler::HS);           // smullHS r1, r5, r6, r7
+    __ smulls(r0, r11, r12, r5, Assembler::MI);        // smullMIs r0, r11, r12, r5
 
 // FourRegNon
-    __ umaal(r8, r9, r2, r5);                          //       umaal r8, r9, r2, r5
-    __ mls(r0, r4, sp, lr, Assembler::EQ);             //       mlsEQ r0, r4, sp, lr
+    __ umaal(r8, r9, r2, r5);                          // umaal r8, r9, r2, r5
+    __ mls(r0, r4, sp, lr, Assembler::EQ);             // mlsEQ r0, r4, sp, lr
 
 // ThreeRegNon
-    __ qadd(r9, r4, sp, Assembler::PL);                //       qaddPL r9, r4, sp
-    __ qsub(r0, r12, r5, Assembler::MI);               //       qsubMI r0, r12, r5
-    __ qdadd(r3, r5, r7);                              //       qdadd r3, r5, r7
-    __ qdsub(r9, r2, r4);                              //       qdsub r9, r2, r4
+    __ qadd(r9, r4, sp, Assembler::PL);                // qaddPL r9, r4, sp
+    __ qsub(r0, r12, r5, Assembler::MI);               // qsubMI r0, r12, r5
+    __ qdadd(r3, r5, r7);                              // qdadd r3, r5, r7
+    __ qdsub(r9, r2, r4);                              // qdsub r9, r2, r4
 
 // FourRegNon
-    __ smlabb(r1, r12, r5, r6);                        //       smlabb r1, r12, r5, r6
-    __ smlabt(r0, r10, r12, r6);                       //       smlabt r0, r10, r12, r6
-    __ smlatb(r8, r1, r3, lr);                         //       smlatb r8, r1, r3, lr
-    __ smlatt(r1, sp, r6, r7);                         //       smlatt r1, sp, r6, r7
-    __ smlawb(r0, r3, r4, r6);                         //       smlawb r0, r3, r4, r6
-    __ smlawt(r11, r4, lr, r7);                        //       smlawt r11, r4, lr, r7
-    __ smlalbb(r0, r10, r6, r7);                       //       smlalbb r0, r10, r6, r7
-    __ smlalbt(r3, r11, r4, lr, Assembler::LS);        //       smlalbtLS r3, r11, r4, lr
-    __ smlaltb(r8, r11, r3, r12);                      //       smlaltb r8, r11, r3, r12
-    __ smlaltt(r8, r1, r3, r5);                        //       smlaltt r8, r1, r3, r5
+    __ smlabb(r1, r12, r5, r6);                        // smlabb r1, r12, r5, r6
+    __ smlabt(r0, r10, r12, r6);                       // smlabt r0, r10, r12, r6
+    __ smlatb(r8, r1, r3, lr);                         // smlatb r8, r1, r3, lr
+    __ smlatt(r1, sp, r6, r7);                         // smlatt r1, sp, r6, r7
+    __ smlawb(r0, r3, r4, r6);                         // smlawb r0, r3, r4, r6
+    __ smlawt(r11, r4, lr, r7);                        // smlawt r11, r4, lr, r7
+    __ smlalbb(r0, r10, r6, r7);                       // smlalbb r0, r10, r6, r7
+    __ smlalbt(r3, r11, r4, lr, Assembler::LS);        // smlalbtLS r3, r11, r4, lr
+    __ smlaltb(r8, r11, r3, r12);                      // smlaltb r8, r11, r3, r12
+    __ smlaltt(r8, r1, r3, r5);                        // smlaltt r8, r1, r3, r5
 
 // ThreeRegNon
-    __ smulwb(r2, r12, sp, Assembler::HS);             //       smulwbHS r2, r12, sp
-    __ smulwt(r8, r12, r6);                            //       smulwt r8, r12, r6
-    __ smulbb(r2, r6, lr, Assembler::GE);              //       smulbbGE r2, r6, lr
-    __ smulbt(r8, r12, r7);                            //       smulbt r8, r12, r7
-    __ smultb(r10, r3, lr, Assembler::EQ);             //       smultbEQ r10, r3, lr
-    __ smultt(r0, r3, sp);                             //       smultt r0, r3, sp
+    __ smulwb(r2, r12, sp, Assembler::HS);             // smulwbHS r2, r12, sp
+    __ smulwt(r8, r12, r6);                            // smulwt r8, r12, r6
+    __ smulbb(r2, r6, lr, Assembler::GE);              // smulbbGE r2, r6, lr
+    __ smulbt(r8, r12, r7);                            // smulbt r8, r12, r7
+    __ smultb(r10, r3, lr, Assembler::EQ);             // smultbEQ r10, r3, lr
+    __ smultt(r0, r3, sp);                             // smultt r0, r3, sp
 
 // MemoryOp
-    __ ldr(r10, Address(r7, r9, lsl(), Address::ADD, Address::post)); //        ldr r10, [r7], r9
-    __ ldrb(r0, Address(r9, 196));                     //       ldrb r0, [r9, #196]
+    __ ldr(r10, Address(r7, r9, lsl(), Address::ADD, Address::post)); // ldr r10, [r7], r9
+    __ ldrb(r0, Address(r9, 196));                     // ldrb r0, [r9, #196]
     __ ldrh(lr, Address(r4, r6, lsl(), Address::ADD, Address::pre)); // ldrh lr, [r4, r6]!
-    __ ldrsb(r6, Address(__ pre(r9, 232)));            //       ldrsb r6, [r9, #232]!
-    __ ldrsh(r2, Address(r1, r1, lsl(), Address::ADD, Address::post)); //       ldrsh r2, [r1], r1
+    __ ldrsb(r6, Address(__ pre(r9, 232)));            // ldrsb r6, [r9, #232]!
+    __ ldrsh(r2, Address(r1, r1, lsl(), Address::ADD, Address::post)); // ldrsh r2, [r1], r1
     __ str(r0, Address(r9, r4, lsl(), Address::ADD, Address::post)); // str r0, [r9], r4
-    __ strb(r3, Address(__ pre(r5, 92)));              //       strb r3, [r5, #92]!
-    __ strh(r2, Address(r8, 160));                     //       strh r2, [r8, #160]
+    __ strb(r3, Address(__ pre(r5, 92)));              // strb r3, [r5, #92]!
+    __ strh(r2, Address(r8, 160));                     // strh r2, [r8, #160]
 
 // MemoryOp
     __ ldr(r8, Address(r12, r8, lsl(), Address::ADD, Address::off)); // ldr r8, [r12, r8]
-    __ ldrb(r11, Address(__ post(r10, 16)));           //       ldrb r11, [r10], #16
-    __ ldrh(r11, Address(r10, r6, lsl(), Address::ADD, Address::off)); //       ldrh r11, [r10, r6]
-    __ ldrsb(r5, Address(r11, r10, lsl(), Address::ADD, Address::pre)); //      ldrsb r5, [r11, r10]!
-    __ ldrsh(r6, Address(r3, r7, lsl(), Address::ADD, Address::off)); //        ldrsh r6, [r3, r7]
-    __ str(r7, Address(sp, r5, lsl(), Address::ADD, Address::pre)); //  str r7, [sp, r5]!
-    __ strb(r2, Address(r10));                         //       strb r2, [r10]
-    __ strh(r6, Address(r4, r3, lsl(), Address::ADD, Address::post)); //        strh r6, [r4], r3
+    __ ldrb(r11, Address(__ post(r10, 16)));           // ldrb r11, [r10], #16
+    __ ldrh(r11, Address(r10, r6, lsl(), Address::ADD, Address::off)); // ldrh r11, [r10, r6]
+    __ ldrsb(r5, Address(r11, r10, lsl(), Address::ADD, Address::pre)); // ldrsb r5, [r11, r10]!
+    __ ldrsh(r6, Address(r3, r7, lsl(), Address::ADD, Address::off)); // ldrsh r6, [r3, r7]
+    __ str(r7, Address(sp, r5, lsl(), Address::ADD, Address::pre)); // str r7, [sp, r5]!
+    __ strb(r2, Address(r10));                         // strb r2, [r10]
+    __ strh(r6, Address(r4, r3, lsl(), Address::ADD, Address::post)); // strh r6, [r4], r3
 
 // MemoryOp
-    __ ldr(r10, Address(r12));                         //       ldr r10, [r12]
-    __ ldrb(r4, Address(__ post(r11, 132)));           //       ldrb r4, [r11], #132
-    __ ldrh(r9, Address(r9, r12, lsl(), Address::ADD, Address::post)); //       ldrh r9, [r9], r12
-    __ ldrsb(r9, Address(__ post(r3, 148)));           //       ldrsb r9, [r3], #148
-    __ ldrsh(r11, Address(__ pre(r2, 148)));           //       ldrsh r11, [r2, #148]!
-    __ str(r11, Address(sp, r11, lsl(), Address::ADD, Address::off)); //        str r11, [sp, r11]
-    __ strb(r1, Address(sp, r10, lsl(), Address::ADD, Address::off)); //        strb r1, [sp, r10]
-    __ strh(r10, Address(lr, r9, lsl(), Address::ADD, Address::post)); //       strh r10, [lr], r9
+    __ ldr(r10, Address(r12));                         // ldr r10, [r12]
+    __ ldrb(r4, Address(__ post(r11, 132)));           // ldrb r4, [r11], #132
+    __ ldrh(r9, Address(r9, r12, lsl(), Address::ADD, Address::post)); // ldrh r9, [r9], r12
+    __ ldrsb(r9, Address(__ post(r3, 148)));           // ldrsb r9, [r3], #148
+    __ ldrsh(r11, Address(__ pre(r2, 148)));           // ldrsh r11, [r2, #148]!
+    __ str(r11, Address(sp, r11, lsl(), Address::ADD, Address::off)); // str r11, [sp, r11]
+    __ strb(r1, Address(sp, r10, lsl(), Address::ADD, Address::off)); // strb r1, [sp, r10]
+    __ strh(r10, Address(lr, r9, lsl(), Address::ADD, Address::post)); // strh r10, [lr], r9
 
 // MemoryOp
-    __ ldr(r6, Address(r3, r4, lsl(), Address::ADD, Address::pre)); //  ldr r6, [r3, r4]!
+    __ ldr(r6, Address(r3, r4, lsl(), Address::ADD, Address::pre)); // ldr r6, [r3, r4]!
     __ ldrb(r4, Address(r6, sp, lsl(), Address::ADD, Address::pre)); // ldrb r4, [r6, sp]!
-    __ ldrh(r6, Address(r7, r10, lsl(), Address::ADD, Address::post)); //       ldrh r6, [r7], r10
-    __ ldrsb(r0, Address(r6, r11, lsl(), Address::ADD, Address::pre)); //       ldrsb r0, [r6, r11]!
-    __ ldrsh(r10, Address(r6, sp, lsl(), Address::ADD, Address::post)); //      ldrsh r10, [r6], sp
+    __ ldrh(r6, Address(r7, r10, lsl(), Address::ADD, Address::post)); // ldrh r6, [r7], r10
+    __ ldrsb(r0, Address(r6, r11, lsl(), Address::ADD, Address::pre)); // ldrsb r0, [r6, r11]!
+    __ ldrsh(r10, Address(r6, sp, lsl(), Address::ADD, Address::post)); // ldrsh r10, [r6], sp
     __ str(r7, Address(r3, r12, lsl(), Address::ADD, Address::off)); // str r7, [r3, r12]
     __ strb(r3, Address(r8, r1, lsl(), Address::ADD, Address::pre)); // strb r3, [r8, r1]!
-    __ strh(r4, Address(r12, 64));                     //       strh r4, [r12, #64]
+    __ strh(r4, Address(r12, 64));                     // strh r4, [r12, #64]
 
     __ bind(near);
 
 // LitMemoryOp
-    __ ldr(r1, near);                                  //       ldr r1, near
-    __ ldrb(r7, __ pc());                              //       ldrb r7, .
-    __ ldrh(r2, near);                                 //       ldrh r2, near
-    __ ldrsb(r10, __ pc());                            //       ldrsb r10, .
-    __ ldrsh(lr, near_post);                           //       ldrsh lr, near_post
+    __ ldr(r1, near);                                  // ldr r1, near
+    __ ldrb(r7, __ pc());                              // ldrb r7, .
+    __ ldrh(r2, near);                                 // ldrh r2, near
+    __ ldrsb(r10, __ pc());                            // ldrsb r10, .
+    __ ldrsh(lr, near_post);                           // ldrsh lr, near_post
 
 // LitMemoryOp
-    __ ldr(r2, __ pc());                               //       ldr r2, .
-    __ ldrb(r3, __ pc());                              //       ldrb r3, .
-    __ ldrh(r7, near_post);                            //       ldrh r7, near_post
-    __ ldrsb(sp, __ pc());                             //       ldrsb sp, .
-    __ ldrsh(r10, near);                               //       ldrsh r10, near
+    __ ldr(r2, __ pc());                               // ldr r2, .
+    __ ldrb(r3, __ pc());                              // ldrb r3, .
+    __ ldrh(r7, near_post);                            // ldrh r7, near_post
+    __ ldrsb(sp, __ pc());                             // ldrsb sp, .
+    __ ldrsh(r10, near);                               // ldrsh r10, near
 
 // LitMemoryOp
-    __ ldr(r5, __ pc());                               //       ldr r5, .
-    __ ldrb(lr, near_post);                            //       ldrb lr, near_post
-    __ ldrh(r5, near_post);                            //       ldrh r5, near_post
-    __ ldrsb(r6, near);                                //       ldrsb r6, near
-    __ ldrsh(r11, near);                               //       ldrsh r11, near
+    __ ldr(r5, __ pc());                               // ldr r5, .
+    __ ldrb(lr, near_post);                            // ldrb lr, near_post
+    __ ldrh(r5, near_post);                            // ldrh r5, near_post
+    __ ldrsb(r6, near);                                // ldrsb r6, near
+    __ ldrsh(r11, near);                               // ldrsh r11, near
 
 // LitMemoryOp
-    __ ldr(r7, near_post);                             //       ldr r7, near_post
-    __ ldrb(r5, near_post);                            //       ldrb r5, near_post
-    __ ldrh(r10, near);                                //       ldrh r10, near
-    __ ldrsb(r6, near_post);                           //       ldrsb r6, near_post
-    __ ldrsh(r9, __ pc());                             //       ldrsh r9, .
+    __ ldr(r7, near_post);                             // ldr r7, near_post
+    __ ldrb(r5, near_post);                            // ldrb r5, near_post
+    __ ldrh(r10, near);                                // ldrh r10, near
+    __ ldrsb(r6, near_post);                           // ldrsb r6, near_post
+    __ ldrsh(r9, __ pc());                             // ldrsh r9, .
 
     __ bind(near_post);
 
 // MemoryRegRegSftOp
-    __ ldr(r0, Address(r0, r10, ::ror(6), Address::ADD, Address::post)); //     ldr r0, [r0], r10, ror #6
-    __ ldrb(r3, Address(r8, lr, ::lsl(9), Address::ADD, Address::off)); //      ldrb r3, [r8, lr, lsl #9]
-    __ str(r5, Address(sp, r3, ::lsl(15), Address::ADD, Address::off)); //      str r5, [sp, r3, lsl #15]
-    __ strb(r9, Address(r9, r5, ::asr(2), Address::ADD, Address::post)); //     strb r9, [r9], r5, asr #2
+    __ ldr(r0, Address(r0, r10, ::ror(6), Address::ADD, Address::post)); // ldr r0, [r0], r10, ror #6
+    __ ldrb(r3, Address(r8, lr, ::lsl(9), Address::ADD, Address::off)); // ldrb r3, [r8, lr, lsl #9]
+    __ str(r5, Address(sp, r3, ::lsl(15), Address::ADD, Address::off)); // str r5, [sp, r3, lsl #15]
+    __ strb(r9, Address(r9, r5, ::asr(2), Address::ADD, Address::post)); // strb r9, [r9], r5, asr #2
 
 // MemoryRegRegSftOp
-    __ ldr(r5, Address(r4, r0, ::ror(6), Address::ADD, Address::off)); //       ldr r5, [r4, r0, ror #6]
-    __ ldrb(lr, Address(r0, r4, ::lsr(9), Address::ADD, Address::off)); //      ldrb lr, [r0, r4, lsr #9]
-    __ str(r5, Address(r12, r12, ::asr(5), Address::ADD, Address::post)); //    str r5, [r12], r12, asr #5
-    __ strb(r3, Address(r1, r7, ::ror(12), Address::ADD, Address::pre)); //     strb r3, [r1, r7, ror #12]!
+    __ ldr(r5, Address(r4, r0, ::ror(6), Address::ADD, Address::off)); // ldr r5, [r4, r0, ror #6]
+    __ ldrb(lr, Address(r0, r4, ::lsr(9), Address::ADD, Address::off)); // ldrb lr, [r0, r4, lsr #9]
+    __ str(r5, Address(r12, r12, ::asr(5), Address::ADD, Address::post)); // str r5, [r12], r12, asr #5
+    __ strb(r3, Address(r1, r7, ::ror(12), Address::ADD, Address::pre)); // strb r3, [r1, r7, ror #12]!
 
 // MemoryRegRegSftOp
-    __ ldr(r6, Address(r2, r3, ::rrx(), Address::ADD, Address::pre)); //        ldr r6, [r2, r3, rrx]!
-    __ ldrb(r8, Address(lr, r2, ::asr(16), Address::ADD, Address::pre)); //     ldrb r8, [lr, r2, asr #16]!
-    __ str(r6, Address(r3, r6, ::ror(7), Address::ADD, Address::pre)); //       str r6, [r3, r6, ror #7]!
-    __ strb(r3, Address(r8, r2, ::lsl(10), Address::ADD, Address::off)); //     strb r3, [r8, r2, lsl #10]
+    __ ldr(r6, Address(r2, r3, ::rrx(), Address::ADD, Address::pre)); // ldr r6, [r2, r3, rrx]!
+    __ ldrb(r8, Address(lr, r2, ::asr(16), Address::ADD, Address::pre)); // ldrb r8, [lr, r2, asr #16]!
+    __ str(r6, Address(r3, r6, ::ror(7), Address::ADD, Address::pre)); // str r6, [r3, r6, ror #7]!
+    __ strb(r3, Address(r8, r2, ::lsl(10), Address::ADD, Address::off)); // strb r3, [r8, r2, lsl #10]
 
 // MemoryRegRegSftOp
-    __ ldr(r11, Address(sp, lr, ::lsl(8), Address::ADD, Address::off)); //      ldr r11, [sp, lr, lsl #8]
-    __ ldrb(r10, Address(sp, r12, ::lsl(4), Address::ADD, Address::pre)); //    ldrb r10, [sp, r12, lsl #4]!
-    __ str(sp, Address(r9, r2, ::asr(2), Address::ADD, Address::off)); //       str sp, [r9, r2, asr #2]
-    __ strb(r7, Address(r11, lr, ::asr(14), Address::ADD, Address::pre)); //    strb r7, [r11, lr, asr #14]!
+    __ ldr(r11, Address(sp, lr, ::lsl(8), Address::ADD, Address::off)); // ldr r11, [sp, lr, lsl #8]
+    __ ldrb(r10, Address(sp, r12, ::lsl(4), Address::ADD, Address::pre)); // ldrb r10, [sp, r12, lsl #4]!
+    __ str(sp, Address(r9, r2, ::asr(2), Address::ADD, Address::off)); // str sp, [r9, r2, asr #2]
+    __ strb(r7, Address(r11, lr, ::asr(14), Address::ADD, Address::pre)); // strb r7, [r11, lr, asr #14]!
 
 // LdStOne
-    __ ldrex(r12, r11);                                //       ldrex r12, [r11]
-    __ ldrexb(r4, r12);                                //       ldrexb r4, [r12]
-    __ ldrexh(r11, r11);                               //       ldrexh r11, [r11]
+    __ ldrex(r12, r11);                                // ldrex r12, [r11]
+    __ ldrexb(r4, r12);                                // ldrexb r4, [r12]
+    __ ldrexh(r11, r11);                               // ldrexh r11, [r11]
 
 // LdStTwo
-    __ strex(r1, r7, lr);                              //       strex r1, r7, [lr]
-    __ strexb(r12, r6, r4);                            //       strexb r12, r6, [r4]
-    __ strexh(r4, r6, r7, Assembler::HS);              //       strexhHS r4, r6, [r7]
+    __ strex(r1, r7, lr);                              // strex r1, r7, [lr]
+    __ strexb(r12, r6, r4);                            // strexb r12, r6, [r4]
+    __ strexh(r4, r6, r7, Assembler::HS);              // strexhHS r4, r6, [r7]
 
 // ThreeRegNon
-    __ sadd16(r3, r4, r7);                             //       sadd16 r3, r4, r7
-    __ sasx(r9, r10, r3, Assembler::AL);               //       sasxAL r9, r10, r3
-    __ ssax(r12, r5, r6);                              //       ssax r12, r5, r6
-    __ ssub16(r12, r5, lr);                            //       ssub16 r12, r5, lr
-    __ sadd8(r0, r10, r7);                             //       sadd8 r0, r10, r7
-    __ ssub8(r0, r8, r2, Assembler::VS);               //       ssub8VS r0, r8, r2
-    __ qadd16(r11, r4, r5, Assembler::PL);             //       qadd16PL r11, r4, r5
-    __ qasx(r11, r3, r12, Assembler::VS);              //       qasxVS r11, r3, r12
-    __ qsax(r0, r3, r5);                               //       qsax r0, r3, r5
-    __ ssub16(r10, r12, r5, Assembler::AL);            //       ssub16AL r10, r12, r5
-    __ qadd8(r10, r6, lr, Assembler::CC);              //       qadd8CC r10, r6, lr
-    __ qsub8(r10, r11, r7);                            //       qsub8 r10, r11, r7
-    __ shadd16(r9, r4, lr, Assembler::PL);             //       shadd16PL r9, r4, lr
-    __ shasx(r1, lr, r7);                              //       shasx r1, lr, r7
-    __ shsax(r9, r11, r5, Assembler::LO);              //       shsaxLO r9, r11, r5
-    __ shsub16(r3, r1, r11, Assembler::GE);            //       shsub16GE r3, r1, r11
-    __ shadd8(sp, r5, r7, Assembler::GT);              //       shadd8GT sp, r5, r7
-    __ shsub8(r1, r5, r7);                             //       shsub8 r1, r5, r7
+    __ sadd16(r3, r4, r7);                             // sadd16 r3, r4, r7
+    __ sasx(r9, r10, r3, Assembler::AL);               // sasxAL r9, r10, r3
+    __ ssax(r12, r5, r6);                              // ssax r12, r5, r6
+    __ ssub16(r12, r5, lr);                            // ssub16 r12, r5, lr
+    __ sadd8(r0, r10, r7);                             // sadd8 r0, r10, r7
+    __ ssub8(r0, r8, r2, Assembler::VS);               // ssub8VS r0, r8, r2
+    __ qadd16(r11, r4, r5, Assembler::PL);             // qadd16PL r11, r4, r5
+    __ qasx(r11, r3, r12, Assembler::VS);              // qasxVS r11, r3, r12
+    __ qsax(r0, r3, r5);                               // qsax r0, r3, r5
+    __ ssub16(r10, r12, r5, Assembler::AL);            // ssub16AL r10, r12, r5
+    __ qadd8(r10, r6, lr, Assembler::CC);              // qadd8CC r10, r6, lr
+    __ qsub8(r10, r11, r7);                            // qsub8 r10, r11, r7
+    __ shadd16(r9, r4, lr, Assembler::PL);             // shadd16PL r9, r4, lr
+    __ shasx(r1, lr, r7);                              // shasx r1, lr, r7
+    __ shsax(r9, r11, r5, Assembler::LO);              // shsaxLO r9, r11, r5
+    __ shsub16(r3, r1, r11, Assembler::GE);            // shsub16GE r3, r1, r11
+    __ shadd8(sp, r5, r7, Assembler::GT);              // shadd8GT sp, r5, r7
+    __ shsub8(r1, r5, r7);                             // shsub8 r1, r5, r7
 
 // ThreeRegNon
-    __ uadd16(r10, r4, r7);                            //       uadd16 r10, r4, r7
-    __ uasx(r1, r9, r7, Assembler::HS);                //       uasxHS r1, r9, r7
-    __ usax(r11, sp, r7);                              //       usax r11, sp, r7
-    __ usub16(r11, r4, lr);                            //       usub16 r11, r4, lr
-    __ uadd8(r2, sp, r7, Assembler::LO);               //       uadd8LO r2, sp, r7
-    __ usub8(r8, r10, lr, Assembler::GT);              //       usub8GT r8, r10, lr
-    __ uqadd16(r3, r12, sp);                           //       uqadd16 r3, r12, sp
-    __ uqasx(r4, sp, r6);                              //       uqasx r4, sp, r6
-    __ uqsax(r1, r10, lr);                             //       uqsax r1, r10, lr
-    __ uqsub16(r2, sp, lr, Assembler::LE);             //       uqsub16LE r2, sp, lr
-    __ uqadd8(r1, r12, r5);                            //       uqadd8 r1, r12, r5
-    __ uqsub8(r0, r4, sp, Assembler::GT);              //       uqsub8GT r0, r4, sp
-    __ uhadd16(r0, r10, r5, Assembler::HI);            //       uhadd16HI r0, r10, r5
-    __ uhasx(r11, r4, r7, Assembler::LE);              //       uhasxLE r11, r4, r7
-    __ uhsax(r1, lr, r9, Assembler::GE);               //       uhsaxGE r1, lr, r9
-    __ uhsub16(r2, r11, lr);                           //       uhsub16 r2, r11, lr
-    __ uhadd8(r9, r4, r5, Assembler::GE);              //       uhadd8GE r9, r4, r5
-    __ uhsub8(r2, sp, lr, Assembler::HI);              //       uhsub8HI r2, sp, lr
+    __ uadd16(r10, r4, r7);                            // uadd16 r10, r4, r7
+    __ uasx(r1, r9, r7, Assembler::HS);                // uasxHS r1, r9, r7
+    __ usax(r11, sp, r7);                              // usax r11, sp, r7
+    __ usub16(r11, r4, lr);                            // usub16 r11, r4, lr
+    __ uadd8(r2, sp, r7, Assembler::LO);               // uadd8LO r2, sp, r7
+    __ usub8(r8, r10, lr, Assembler::GT);              // usub8GT r8, r10, lr
+    __ uqadd16(r3, r12, sp);                           // uqadd16 r3, r12, sp
+    __ uqasx(r4, sp, r6);                              // uqasx r4, sp, r6
+    __ uqsax(r1, r10, lr);                             // uqsax r1, r10, lr
+    __ uqsub16(r2, sp, lr, Assembler::LE);             // uqsub16LE r2, sp, lr
+    __ uqadd8(r1, r12, r5);                            // uqadd8 r1, r12, r5
+    __ uqsub8(r0, r4, sp, Assembler::GT);              // uqsub8GT r0, r4, sp
+    __ uhadd16(r0, r10, r5, Assembler::HI);            // uhadd16HI r0, r10, r5
+    __ uhasx(r11, r4, r7, Assembler::LE);              // uhasxLE r11, r4, r7
+    __ uhsax(r1, lr, r9, Assembler::GE);               // uhsaxGE r1, lr, r9
+    __ uhsub16(r2, r11, lr);                           // uhsub16 r2, r11, lr
+    __ uhadd8(r9, r4, r5, Assembler::GE);              // uhadd8GE r9, r4, r5
+    __ uhsub8(r2, sp, lr, Assembler::HI);              // uhsub8HI r2, sp, lr
 
 // PKUPSATREV
-    __ sxtab16(r10, r3, r7, ::ror(16));                //       sxtab16 r10, r3, r7, ROR #16
-    __ sxtab(r9, r5, r7, ::ror(24), Assembler::CS);    //       sxtabCS r9, r5, r7, ROR #24
-    __ sxtah(r3, r5, r7, ::ror(8));                    //       sxtah r3, r5, r7, ROR #8
-    __ uxtab16(r8, r4, r6, ::ror(8), Assembler::AL);   //       uxtab16AL r8, r4, r6, ROR #8
-    __ uxtab(r0, r11, sp, ::rrx(), Assembler::EQ);     //       uxtabEQ r0, r11, sp, ROR #0
-    __ uxtah(r9, r12, r5, ::rrx());                    //       uxtah r9, r12, r5, ROR #0
+    __ sxtab16(r10, r3, r7, ::ror(16));                // sxtab16 r10, r3, r7, ROR #16
+    __ sxtab(r9, r5, r7, ::ror(24), Assembler::CS);    // sxtabCS r9, r5, r7, ROR #24
+    __ sxtah(r3, r5, r7, ::ror(8));                    // sxtah r3, r5, r7, ROR #8
+    __ uxtab16(r8, r4, r6, ::ror(8), Assembler::AL);   // uxtab16AL r8, r4, r6, ROR #8
+    __ uxtab(r0, r11, sp, ::rrx(), Assembler::EQ);     // uxtabEQ r0, r11, sp, ROR #0
+    __ uxtah(r9, r12, r5, ::rrx());                    // uxtah r9, r12, r5, ROR #0
 
 // PKUPSATREV
-    __ sxtb16(r3, r11, ::ror(16), Assembler::GE);      //       sxtb16GE r3, r11, ROR #16
-    __ sxtb(r2, r6, ::rrx(), Assembler::HI);           //       sxtbHI r2, r6, ROR #0
-    __ sxth(r3, sp, ::ror(24), Assembler::GT);         //       sxthGT r3, sp, ROR #24
-    __ uxtb16(r12, r5, ::ror(16));                     //       uxtb16 r12, r5, ROR #16
-    __ uxtb(r12, r5, ::ror(16));                       //       uxtb r12, r5, ROR #16
-    __ uxth(r8, r5, ::ror(16));                        //       uxth r8, r5, ROR #16
+    __ sxtb16(r3, r11, ::ror(16), Assembler::GE);      // sxtb16GE r3, r11, ROR #16
+    __ sxtb(r2, r6, ::rrx(), Assembler::HI);           // sxtbHI r2, r6, ROR #0
+    __ sxth(r3, sp, ::ror(24), Assembler::GT);         // sxthGT r3, sp, ROR #24
+    __ uxtb16(r12, r5, ::ror(16));                     // uxtb16 r12, r5, ROR #16
+    __ uxtb(r12, r5, ::ror(16));                       // uxtb r12, r5, ROR #16
+    __ uxth(r8, r5, ::ror(16));                        // uxth r8, r5, ROR #16
 
 // TwoRegNon
-    __ rev(r10, r4, Assembler::EQ);                    //       revEQ r10, r4
-    __ rev16(r8, r12, Assembler::GE);                  //       rev16GE r8, r12
-    __ rbit(lr, r7);                                   //       rbit lr, r7
-    __ revsh(sp, r7, Assembler::GT);                   //       revshGT sp, r7
+    __ rev(r10, r4, Assembler::EQ);                    // revEQ r10, r4
+    __ rev16(r8, r12, Assembler::GE);                  // rev16GE r8, r12
+    __ rbit(lr, r7);                                   // rbit lr, r7
+    __ revsh(sp, r7, Assembler::GT);                   // revshGT sp, r7
 
 // ThreeRegNon
-    __ sdiv(r9, sp, lr);                               //       sdiv r9, sp, lr
-    __ udiv(r2, r12, r6);                              //       udiv r2, r12, r6
+    __ sdiv(r9, sp, lr);                               // sdiv r9, sp, lr
+    __ udiv(r2, r12, r6);                              // udiv r2, r12, r6
 
 // TwoRegTwoImm
-    __ sbfx(r0, r1, (unsigned)20U, (unsigned)3U, Assembler::MI); //     sbfxMI r0, r1, #20, #3
-    __ ubfx(r9, r2, (unsigned)16U, (unsigned)15U);     //       ubfx r9, r2, #16, #15
-    __ bfi(r1, r11, (unsigned)27U, (unsigned)3U, Assembler::HI); //     bfiHI r1, r11, #27, #3
+    __ sbfx(r0, r1, (unsigned)20U, (unsigned)3U, Assembler::MI); // sbfxMI r0, r1, #20, #3
+    __ ubfx(r9, r2, (unsigned)16U, (unsigned)15U);     // ubfx r9, r2, #16, #15
+    __ bfi(r1, r11, (unsigned)27U, (unsigned)3U, Assembler::HI); // bfiHI r1, r11, #27, #3
 
 // TwoRegTwoImm
-    __ bfc(r3, (unsigned)7U, (unsigned)10U);           //       bfc r3, #7, #10
+    __ bfc(r3, (unsigned)7U, (unsigned)10U);           // bfc r3, #7, #10
 
 // MultipleMemOp
-    __ stmda(r6, 3435U, false);                        //       stmda r6, {r0, r1, r3, r5, r6, r8, r10, r11}
-    __ stmed(r4, 14559U, false);                       //       stmed r4, {r0, r1, r2, r3, r4, r6, r7, r11, r12, sp}
-    __ ldmda(r0, 57812U, false);                       //       ldmda r0, {r2, r4, r6, r7, r8, sp, lr, pc}
-    __ ldmfa(r12, 39027U, true);                       //       ldmfa r12!, {r0, r1, r4, r5, r6, r11, r12, pc}
-    __ stmia(r9, 12733U, true);                        //       stmia r9!, {r0, r2, r3, r4, r5, r7, r8, r12, sp}
-    __ stmea(r11, 21955U, false);                      //       stmea r11, {r0, r1, r6, r7, r8, r10, r12, lr}
-    __ ldmia(r12, 48418U, true);                       //       ldmia r12!, {r1, r5, r8, r10, r11, r12, sp, pc}
-    __ ldmfd(sp, 41226U, true);                        //       ldmfd sp!, {r1, r3, r8, sp, pc}
-    __ stmdb(r11, 8729U, true);                        //       stmdb r11!, {r0, r3, r4, r9, sp}
-    __ stmfd(r9, 36309U, true);                        //       stmfd r9!, {r0, r2, r4, r6, r7, r8, r10, r11, pc}
-    __ ldmdb(r5, 24667U, true);                        //       ldmdb r5!, {r0, r1, r3, r4, r6, sp, lr}
-    __ ldmea(r1, 37287U, false);                       //       ldmea r1, {r0, r1, r2, r5, r7, r8, r12, pc}
-    __ stmib(r11, 28266U, true);                       //       stmib r11!, {r1, r3, r5, r6, r9, r10, r11, sp, lr}
-    __ stmfa(r11, 17671U, false);                      //       stmfa r11, {r0, r1, r2, r8, r10, lr}
-    __ ldmib(r0, 21452U, true);                        //       ldmib r0!, {r2, r3, r6, r7, r8, r9, r12, lr}
-    __ ldmed(r1, 11751U, false);                       //       ldmed r1, {r0, r1, r2, r5, r6, r7, r8, r10, r11, sp}
+    __ stmda(r6, 3435U, false);                        // stmda r6, {r0, r1, r3, r5, r6, r8, r10, r11}
+    __ stmed(r4, 14559U, false);                       // stmed r4, {r0, r1, r2, r3, r4, r6, r7, r11, r12, sp}
+    __ ldmda(r0, 57812U, false);                       // ldmda r0, {r2, r4, r6, r7, r8, sp, lr, pc}
+    __ ldmfa(r12, 39027U, true);                       // ldmfa r12!, {r0, r1, r4, r5, r6, r11, r12, pc}
+    __ stmia(r9, 12733U, true);                        // stmia r9!, {r0, r2, r3, r4, r5, r7, r8, r12, sp}
+    __ stmea(r11, 21955U, false);                      // stmea r11, {r0, r1, r6, r7, r8, r10, r12, lr}
+    __ ldmia(r12, 48418U, true);                       // ldmia r12!, {r1, r5, r8, r10, r11, r12, sp, pc}
+    __ ldmfd(sp, 41226U, true);                        // ldmfd sp!, {r1, r3, r8, sp, pc}
+    __ stmdb(r11, 8729U, true);                        // stmdb r11!, {r0, r3, r4, r9, sp}
+    __ stmfd(r9, 36309U, true);                        // stmfd r9!, {r0, r2, r4, r6, r7, r8, r10, r11, pc}
+    __ ldmdb(r5, 24667U, true);                        // ldmdb r5!, {r0, r1, r3, r4, r6, sp, lr}
+    __ ldmea(r1, 37287U, false);                       // ldmea r1, {r0, r1, r2, r5, r7, r8, r12, pc}
+    __ stmib(r11, 28266U, true);                       // stmib r11!, {r1, r3, r5, r6, r9, r10, r11, sp, lr}
+    __ stmfa(r11, 17671U, false);                      // stmfa r11, {r0, r1, r2, r8, r10, lr}
+    __ ldmib(r0, 21452U, true);                        // ldmib r0!, {r2, r3, r6, r7, r8, r9, r12, lr}
+    __ ldmed(r1, 11751U, false);                       // ldmed r1, {r0, r1, r2, r5, r6, r7, r8, r10, r11, sp}
 
 // BranchLabel
-    __ b(forth, Assembler::CS);                        //       bCS forth
-    __ bl(__ pc(), Assembler::MI);                     //       blMI .
+    __ b(forth, Assembler::CS);                        // bCS forth
+    __ bl(__ pc(), Assembler::MI);                     // blMI .
 
 // OneRegNon
-    __ b(r0, Assembler::VS);                           //       bxVS r0
-    __ bl(r3);                                         //       blx r3
+    __ b(r0, Assembler::VS);                           // bxVS r0
+    __ bl(r3);                                         // blx r3
 
 // BranchLabel
-    __ b(__ pc(), Assembler::AL);                      //       bAL .
-    __ bl(__ pc());                                    //       bl .
+    __ b(__ pc(), Assembler::AL);                      // bAL .
+    __ bl(__ pc());                                    // bl .
 
 // OneRegNon
-    __ b(r0, Assembler::VS);                           //       bxVS r0
-    __ bl(r5);                                         //       blx r5
+    __ b(r0, Assembler::VS);                           // bxVS r0
+    __ bl(r5);                                         // blx r5
 
 // BranchLabel
-    __ b(forth, Assembler::LE);                        //       bLE forth
-    __ bl(__ pc(), Assembler::MI);                     //       blMI .
+    __ b(forth, Assembler::LE);                        // bLE forth
+    __ bl(__ pc(), Assembler::MI);                     // blMI .
 
 // OneRegNon
-    __ b(r9, Assembler::NE);                           //       bxNE r9
-    __ bl(r12);                                        //       blx r12
+    __ b(r9, Assembler::NE);                           // bxNE r9
+    __ bl(r12);                                        // blx r12
 
 // BranchLabel
-    __ b(back);                                        //       b back
-    __ bl(__ pc(), Assembler::HI);                     //       blHI .
+    __ b(back);                                        // b back
+    __ bl(__ pc(), Assembler::HI);                     // blHI .
 
 // OneRegNon
-    __ b(r1, Assembler::VC);                           //       bxVC r1
-    __ bl(r7, Assembler::GT);                          //       blxGT r7
+    __ b(r1, Assembler::VC);                           // bxVC r1
+    __ bl(r7, Assembler::GT);                          // blxGT r7
 
 // BranchLabel
-    __ b(back, Assembler::GE);                         //       bGE back
-    __ bl(__ pc(), Assembler::HI);                     //       blHI .
+    __ b(back, Assembler::GE);                         // bGE back
+    __ bl(__ pc(), Assembler::HI);                     // blHI .
 
 // OneRegNon
-    __ b(r12);                                         //       bx r12
-    __ bl(r7, Assembler::CC);                          //       blxCC r7
+    __ b(r12);                                         // bx r12
+    __ bl(r7, Assembler::CC);                          // blxCC r7
 
 // BranchLabel
-    __ b(__ pc());                                     //       b .
-    __ bl(back, Assembler::GT);                        //       blGT back
+    __ b(__ pc());                                     // b .
+    __ bl(back, Assembler::GT);                        // blGT back
 
 // OneRegNon
-    __ b(r1, Assembler::GE);                           //       bxGE r1
-    __ bl(r0);                                         //       blx r0
+    __ b(r1, Assembler::GE);                           // bxGE r1
+    __ bl(r0);                                         // blx r0
 
 // BranchLabel
-    __ b(__ pc());                                     //       b .
-    __ bl(forth);                                      //       bl forth
+    __ b(__ pc());                                     // b .
+    __ bl(forth);                                      // bl forth
 
 // OneRegNon
-    __ b(lr, Assembler::GT);                           //       bxGT lr
-    __ bl(r11, Assembler::NE);                         //       blxNE r11
+    __ b(lr, Assembler::GT);                           // bxGT lr
+    __ bl(r11, Assembler::NE);                         // blxNE r11
 
 // BranchLabel
-    __ b(__ pc(), Assembler::CS);                      //       bCS .
-    __ bl(__ pc());                                    //       bl .
+    __ b(__ pc(), Assembler::CS);                      // bCS .
+    __ bl(__ pc());                                    // bl .
 
 // OneRegNon
-    __ b(r10, Assembler::HS);                          //       bxHS r10
-    __ bl(r4);                                         //       blx r4
+    __ b(r10, Assembler::HS);                          // bxHS r10
+    __ bl(r4);                                         // blx r4
 
 // BranchLabel
-    __ b(back, Assembler::AL);                         //       bAL back
-    __ bl(__ pc());                                    //       bl .
+    __ b(back, Assembler::AL);                         // bAL back
+    __ bl(__ pc());                                    // bl .
 
 // OneRegNon
-    __ b(r12, Assembler::LO);                          //       bxLO r12
-    __ bl(r8);                                         //       blx r8
+    __ b(r12, Assembler::LO);                          // bxLO r12
+    __ bl(r8);                                         // blx r8
 
 // BranchLabel
-    __ b(forth);                                       //       b forth
-    __ bl(__ pc());                                    //       bl .
+    __ b(forth);                                       // b forth
+    __ bl(__ pc());                                    // bl .
 
 // OneRegNon
-    __ b(r10);                                         //       bx r10
-    __ bl(r1);                                         //       blx r1
+    __ b(r10);                                         // bx r10
+    __ bl(r1);                                         // blx r1
 
 // ThreeFltNon
-    __ vmla_f32(d2, d4, d6, Assembler::MI);            //       vmlaMI.f32 s4, s8, s12
-    __ vmls_f32(d2, d5, d5);                           //       vmls.f32 s4, s10, s10
-    __ vnmla_f32(d1, d5, d6);                          //       vnmla.f32 s2, s10, s12
-    __ vnmls_f32(d4, d3, d4, Assembler::LT);           //       vnmlsLT.f32 s8, s6, s8
-    __ vnmul_f32(d3, d6, d7, Assembler::MI);           //       vnmulMI.f32 s6, s12, s14
-    __ vadd_f32(d0, d1, d0);                           //       vadd.f32 s0, s2, s0
-    __ vsub_f32(d1, d2, d5, Assembler::AL);            //       vsubAL.f32 s2, s4, s10
-    __ vdiv_f32(d0, d1, d6, Assembler::CS);            //       vdivCS.f32 s0, s2, s12
+    __ vmla_f32(d2, d4, d6, Assembler::MI);            // vmlaMI.f32 s4, s8, s12
+    __ vmls_f32(d2, d5, d5);                           // vmls.f32 s4, s10, s10
+    __ vnmla_f32(d1, d5, d6);                          // vnmla.f32 s2, s10, s12
+    __ vnmls_f32(d4, d3, d4, Assembler::LT);           // vnmlsLT.f32 s8, s6, s8
+    __ vnmul_f32(d3, d6, d7, Assembler::MI);           // vnmulMI.f32 s6, s12, s14
+    __ vadd_f32(d0, d1, d0);                           // vadd.f32 s0, s2, s0
+    __ vsub_f32(d1, d2, d5, Assembler::AL);            // vsubAL.f32 s2, s4, s10
+    __ vdiv_f32(d0, d1, d6, Assembler::CS);            // vdivCS.f32 s0, s2, s12
 
 // ThreeFltNon
-    __ vmla_f64(d0, d3, d6);                           //       vmla.f64 d0, d3, d6
-    __ vmls_f64(d0, d1, d5);                           //       vmls.f64 d0, d1, d5
-    __ vnmla_f64(d1, d4, d6);                          //       vnmla.f64 d1, d4, d6
-    __ vnmls_f64(d0, d1, d1, Assembler::NE);           //       vnmlsNE.f64 d0, d1, d1
-    __ vnmul_f64(d3, d5, d5, Assembler::NE);           //       vnmulNE.f64 d3, d5, d5
-    __ vadd_f64(d0, d2, d4, Assembler::LO);            //       vaddLO.f64 d0, d2, d4
-    __ vsub_f64(d1, d2, d4);                           //       vsub.f64 d1, d2, d4
-    __ vdiv_f64(d0, d1, d5, Assembler::MI);            //       vdivMI.f64 d0, d1, d5
+    __ vmla_f64(d0, d3, d6);                           // vmla.f64 d0, d3, d6
+    __ vmls_f64(d0, d1, d5);                           // vmls.f64 d0, d1, d5
+    __ vnmla_f64(d1, d4, d6);                          // vnmla.f64 d1, d4, d6
+    __ vnmls_f64(d0, d1, d1, Assembler::NE);           // vnmlsNE.f64 d0, d1, d1
+    __ vnmul_f64(d3, d5, d5, Assembler::NE);           // vnmulNE.f64 d3, d5, d5
+    __ vadd_f64(d0, d2, d4, Assembler::LO);            // vaddLO.f64 d0, d2, d4
+    __ vsub_f64(d1, d2, d4);                           // vsub.f64 d1, d2, d4
+    __ vdiv_f64(d0, d1, d5, Assembler::MI);            // vdivMI.f64 d0, d1, d5
 
 // TwoFltNon
-    __ vabs_f32(d3, d3);                               //       vabs.f32 s6, s6
-    __ vneg_f32(d3, d4, Assembler::PL);                //       vnegPL.f32 s6, s8
-    __ vsqrt_f32(d0, d4);                              //       vsqrt.f32 s0, s8
+    __ vabs_f32(d3, d3);                               // vabs.f32 s6, s6
+    __ vneg_f32(d3, d4, Assembler::PL);                // vnegPL.f32 s6, s8
+    __ vsqrt_f32(d0, d4);                              // vsqrt.f32 s0, s8
 
 // TwoFltNon
-    __ vabs_f64(d0, d4);                               //       vabs.f64 d0, d4
-    __ vneg_f64(d1, d4);                               //       vneg.f64 d1, d4
-    __ vsqrt_f64(d0, d1);                              //       vsqrt.f64 d0, d1
+    __ vabs_f64(d0, d4);                               // vabs.f64 d0, d4
+    __ vneg_f64(d1, d4);                               // vneg.f64 d1, d4
+    __ vsqrt_f64(d0, d1);                              // vsqrt.f64 d0, d1
 
 // vmov_f32
-    __ vmov_f32(d0, lr, Assembler::PL);                //       vmovPL.f32 s0, lr
+    __ vmov_f32(d0, lr, Assembler::PL);                // vmovPL.f32 s0, lr
 
 // vmov_f32
-    __ vmov_f32(r11, d4);                              //       vmov.f32 r11, s8
+    __ vmov_f32(r11, d4);                              // vmov.f32 r11, s8
 
 // vmov_f64
-    __ vmov_f64(d1, r11, lr, Assembler::LT);           //       vmovLT.f64 d1, r11, lr
+    __ vmov_f64(d1, r11, lr, Assembler::LT);           // vmovLT.f64 d1, r11, lr
 
 // vmov_f64
-    __ vmov_f64(r7, r5, d5);                           //       vmov.f64 r7, r5, d5
+    __ vmov_f64(r7, r5, d5);                           // vmov.f64 r7, r5, d5
 
 // vmov_f32
-    __ vmov_f32(d4, d6);                               //       vmov.f32 s8, s12
+    __ vmov_f32(d4, d6);                               // vmov.f32 s8, s12
 
 // vmov_f64
-    __ vmov_f64(d1, d2, Assembler::HI);                //       vmovHI.f64 d1, d2
+    __ vmov_f64(d1, d2, Assembler::HI);                // vmovHI.f64 d1, d2
 
 // vmov_f32
-    __ vmov_f32(d2, 1.0f, Assembler::VS);              //       vmovVS.f32 s4, #1.0
+    __ vmov_f32(d2, 1.0f, Assembler::VS);              // vmovVS.f32 s4, #1.0
 
 // vmov_f64
-    __ vmov_f64(d2, 1.0);                              //       vmov.f64 d2, #1.0
+    __ vmov_f64(d2, 1.0);                              // vmov.f64 d2, #1.0
 
 // vmov_f32
-    __ vmov_f32(d3, 2.0f);                             //       vmov.f32 s6, #2.0
+    __ vmov_f32(d3, 2.0f);                             // vmov.f32 s6, #2.0
 
 // vmov_f64
-    __ vmov_f64(d1, 2.0);                              //       vmov.f64 d1, #2.0
+    __ vmov_f64(d1, 2.0);                              // vmov.f64 d1, #2.0
 
 // vector memory
-    __ vldr_f32(d2, Address(r5, 116));                 //       vldr.f32 s4, [r5, #116]
-    __ vstr_f32(d1, Address(r1, 56), Assembler::CC);   //       vstrCC.f32 s2, [r1, #56]
+    __ vldr_f32(d2, Address(r5, 116));                 // vldr.f32 s4, [r5, #116]
+    __ vstr_f32(d1, Address(r1, 56), Assembler::CC);   // vstrCC.f32 s2, [r1, #56]
 
 // vector memory
-    __ vldr_f64(d7, Address(r5, 16), Assembler::NE);   //       vldrNE.f64 d7, [r5, #16]
-    __ vstr_f64(d6, Address(r1, 228));                 //       vstr.f64 d6, [r1, #228]
+    __ vldr_f64(d7, Address(r5, 16), Assembler::NE);   // vldrNE.f64 d7, [r5, #16]
+    __ vstr_f64(d6, Address(r1, 228));                 // vstr.f64 d6, [r1, #228]
 
     __ bind(near_flt);
 
 // vector memory
-    __ vldr_f32(d1, near_post_flt);                    //       vldr.f32 s2, near_post_flt
-    __ vstr_f32(d3, near_post_flt);                    //       vstr.f32 s6, near_post_flt
+    __ vldr_f32(d1, near_post_flt);                    // vldr.f32 s2, near_post_flt
+    __ vstr_f32(d3, near_post_flt);                    // vstr.f32 s6, near_post_flt
 
 // vector memory
-    __ vldr_f64(d2, near_flt, Assembler::LT);          //       vldrLT.f64 d2, near_flt
-    __ vstr_f64(d3, __ pc(), Assembler::GT);           //       vstrGT.f64 d3, .
+    __ vldr_f64(d2, near_flt, Assembler::LT);          // vldrLT.f64 d2, near_flt
+    __ vstr_f64(d3, __ pc(), Assembler::GT);           // vstrGT.f64 d3, .
 
 // vector memory
-    __ vldr_f32(d2, near_post_flt, Assembler::CC);     //       vldrCC.f32 s4, near_post_flt
-    __ vstr_f32(d0, near_post_flt);                    //       vstr.f32 s0, near_post_flt
+    __ vldr_f32(d2, near_post_flt, Assembler::CC);     // vldrCC.f32 s4, near_post_flt
+    __ vstr_f32(d0, near_post_flt);                    // vstr.f32 s0, near_post_flt
 
 // vector memory
-    __ vldr_f64(d4, near_post_flt, Assembler::GT);     //       vldrGT.f64 d4, near_post_flt
-    __ vstr_f64(d0, near_flt);                         //       vstr.f64 d0, near_flt
+    __ vldr_f64(d4, near_post_flt, Assembler::GT);     // vldrGT.f64 d4, near_post_flt
+    __ vstr_f64(d0, near_flt);                         // vstr.f64 d0, near_flt
 
 // vector memory
-    __ vldr_f32(d4, near_post_flt);                    //       vldr.f32 s8, near_post_flt
-    __ vstr_f32(d3, near_post_flt);                    //       vstr.f32 s6, near_post_flt
+    __ vldr_f32(d4, near_post_flt);                    // vldr.f32 s8, near_post_flt
+    __ vstr_f32(d3, near_post_flt);                    // vstr.f32 s6, near_post_flt
 
 // vector memory
-    __ vldr_f64(d4, near_flt, Assembler::PL);          //       vldrPL.f64 d4, near_flt
-    __ vstr_f64(d5, near_flt);                         //       vstr.f64 d5, near_flt
+    __ vldr_f64(d4, near_flt, Assembler::PL);          // vldrPL.f64 d4, near_flt
+    __ vstr_f64(d5, near_flt);                         // vstr.f64 d5, near_flt
 
 // vector memory
-    __ vldr_f32(d4, near_post_flt, Assembler::LS);     //       vldrLS.f32 s8, near_post_flt
-    __ vstr_f32(d6, __ pc(), Assembler::CC);           //       vstrCC.f32 s12, .
+    __ vldr_f32(d4, near_post_flt, Assembler::LS);     // vldrLS.f32 s8, near_post_flt
+    __ vstr_f32(d6, __ pc(), Assembler::CC);           // vstrCC.f32 s12, .
 
 // vector memory
-    __ vldr_f64(d6, near_post_flt, Assembler::AL);     //       vldrAL.f64 d6, near_post_flt
-    __ vstr_f64(d1, near_post_flt, Assembler::LT);     //       vstrLT.f64 d1, near_post_flt
+    __ vldr_f64(d6, near_post_flt, Assembler::AL);     // vldrAL.f64 d6, near_post_flt
+    __ vstr_f64(d1, near_post_flt, Assembler::LT);     // vstrLT.f64 d1, near_post_flt
 
     __ bind(near_post_flt);
 
 // FltMultMemOp
-    __ vldmia_f32(r1, 4U, false);                      //       vldmia.f32 r1, {s4}
-    __ vstmia_f32(r6, 4U, true, Assembler::CS);        //       vstmiaCS.f32 r6!, {s4}
+    __ vldmia_f32(r1, 4U, false);                      // vldmia.f32 r1, {s4}
+    __ vstmia_f32(r6, 4U, true, Assembler::CS);        // vstmiaCS.f32 r6!, {s4}
 
 // DblMultMemOp
-    __ vldmia_f64(r9, 30U, true);                      //       vldmia.f64 r9!, {d1, d2, d3, d4}
-    __ vstmia_f64(r3, 192U, true);                     //       vstmia.f64 r3!, {d6, d7}
+    __ vldmia_f64(r9, 30U, true);                      // vldmia.f64 r9!, {d1, d2, d3, d4}
+    __ vstmia_f64(r3, 192U, true);                     // vstmia.f64 r3!, {d6, d7}
 
 // FltMultMemOp
-    __ vldmdb_f32(r2, 8U, Assembler::VS);              //       vldmdbVS.f32 r2!, {s6}
-    __ vstmdb_f32(r6, 128U);                           //       vstmdb.f32 r6!, {s14}
+    __ vldmdb_f32(r2, 8U, Assembler::VS);              // vldmdbVS.f32 r2!, {s6}
+    __ vstmdb_f32(r6, 128U);                           // vstmdb.f32 r6!, {s14}
 
 // DblMultMemOp
-    __ vldmdb_f64(sp, 240U);                           //       vldmdb.f64 sp!, {d4, d5, d6, d7}
-    __ vstmdb_f64(r0, 224U);                           //       vstmdb.f64 r0!, {d5, d6, d7}
+    __ vldmdb_f64(sp, 240U);                           // vldmdb.f64 sp!, {d4, d5, d6, d7}
+    __ vstmdb_f64(r0, 224U);                           // vstmdb.f64 r0!, {d5, d6, d7}
 
 // vcmp_f32
-    __ vcmp_f32(d1, d1);                               //       vcmp.f32 s2, s2
+    __ vcmp_f32(d1, d1);                               // vcmp.f32 s2, s2
 
 // vcmpe_f32
-    __ vcmpe_f32(d4, d4, Assembler::VC);               //       vcmpeVC.f32 s8, s8
+    __ vcmpe_f32(d4, d4, Assembler::VC);               // vcmpeVC.f32 s8, s8
 
 // vcmp_f64
-    __ vcmp_f64(d0, d6);                               //       vcmp.f64 d0, d6
+    __ vcmp_f64(d0, d6);                               // vcmp.f64 d0, d6
 
 // vcmpe_f64
-    __ vcmpe_f64(d3, d7, Assembler::GE);               //       vcmpeGE.f64 d3, d7
+    __ vcmpe_f64(d3, d7, Assembler::GE);               // vcmpeGE.f64 d3, d7
 
 // vcmp_f32
-    __ vcmp_f32(d1, 0.0f, Assembler::LT);              //       vcmpLT.f32 s2, #0.0
+    __ vcmp_f32(d1, 0.0f, Assembler::LT);              // vcmpLT.f32 s2, #0.0
 
 // vcmpe_f32
-    __ vcmpe_f32(d7, 0.0f, Assembler::GT);             //       vcmpeGT.f32 s14, #0.0
+    __ vcmpe_f32(d7, 0.0f, Assembler::GT);             // vcmpeGT.f32 s14, #0.0
 
 // vcmp_f64
-    __ vcmp_f64(d4, 0.0);                              //       vcmp.f64 d4, #0.0
+    __ vcmp_f64(d4, 0.0);                              // vcmp.f64 d4, #0.0
 
 // vcmpe_f64
-    __ vcmpe_f64(d1, 0.0);                             //       vcmpe.f64 d1, #0.0
+    __ vcmpe_f64(d1, 0.0);                             // vcmpe.f64 d1, #0.0
 
 // vcvt
-    __ vcvt_s32_f32(d1, d3, Assembler::VS);            //       vcvtVS.s32.f32 s2, s6
-    __ vcvt_u32_f32(d3, d7, Assembler::GT);            //       vcvtGT.u32.f32 s6, s14
-    __ vcvt_f32_s32(d0, d1, Assembler::CC);            //       vcvtCC.f32.s32 s0, s2
-    __ vcvt_f32_u32(d1, d2, Assembler::CC);            //       vcvtCC.f32.u32 s2, s4
+    __ vcvt_s32_f32(d1, d3, Assembler::VS);            // vcvtVS.s32.f32 s2, s6
+    __ vcvt_u32_f32(d3, d7, Assembler::GT);            // vcvtGT.u32.f32 s6, s14
+    __ vcvt_f32_s32(d0, d1, Assembler::CC);            // vcvtCC.f32.s32 s0, s2
+    __ vcvt_f32_u32(d1, d2, Assembler::CC);            // vcvtCC.f32.u32 s2, s4
 
 // vcvt
-    __ vcvt_s32_f64(d2, d4, Assembler::HI);            //       vcvtHI.s32.f64 s4, d4
-    __ vcvt_u32_f64(d3, d6, Assembler::HI);            //       vcvtHI.u32.f64 s6, d6
-    __ vcvt_f32_f64(d3, d7, Assembler::LS);            //       vcvtLS.f32.f64 s6, d7
+    __ vcvt_s32_f64(d2, d4, Assembler::HI);            // vcvtHI.s32.f64 s4, d4
+    __ vcvt_u32_f64(d3, d6, Assembler::HI);            // vcvtHI.u32.f64 s6, d6
+    __ vcvt_f32_f64(d3, d7, Assembler::LS);            // vcvtLS.f32.f64 s6, d7
 
 // vcvt
-    __ vcvt_f64_s32(d3, d4);                           //       vcvt.f64.s32 d3, s8
-    __ vcvt_f64_u32(d5, d7, Assembler::EQ);            //       vcvtEQ.f64.u32 d5, s14
-    __ vcvt_f64_f32(d4, d5, Assembler::AL);            //       vcvtAL.f64.f32 d4, s10
+    __ vcvt_f64_s32(d3, d4);                           // vcvt.f64.s32 d3, s8
+    __ vcvt_f64_u32(d5, d7, Assembler::EQ);            // vcvtEQ.f64.u32 d5, s14
+    __ vcvt_f64_f32(d4, d5, Assembler::AL);            // vcvtAL.f64.f32 d4, s10
 
 // BKPT
-    __ bkpt((unsigned)26U);                            //       bkpt #26
+    __ bkpt((unsigned)26U);                            // bkpt #26
 
     __ bind(forth);
 
@@ -1603,8 +1603,8 @@
 //This should really be in the macroassembler
 void Assembler::mov_immediate32(Register dst, u_int32_t imm32, Condition cond, bool s)
 {
-        // Need to move a full 32 bit immediate, for example if we're loading an address that
-        // might change later and therefore need to be updated.
+  // Need to move a full 32 bit immediate, for example if we're loading an address that
+  // might change later and therefore need to be updated.
   if (VM_Version::features() & (FT_ARMV7 | FT_ARMV6T2))  {
     //Use a movw and a movt
     Assembler::movw_i(dst, (unsigned)(imm32 & 0xffff), cond);
@@ -1614,13 +1614,13 @@
       Assembler::cmp(dst, 0);
     }
   } else {
-        // Sadly we don't have movw, movt
-        // instead emit a mov and three orr
-        mov_i(dst,    imm32 & (0xff      ), cond);
-        orr(dst, dst, imm32 & (0xff << 8 ), cond);
-        orr(dst, dst, imm32 & (0xff << 16), cond);
-        if(s) orrs(dst, dst, imm32 & (0xff << 24), cond);
-        else  orr (dst, dst, imm32 & (0xff << 24), cond);
+    // Sadly we don't have movw, movt
+    // instead emit a mov and three orr
+    mov_i(dst,    imm32 & (0xff      ), cond);
+    orr(dst, dst, imm32 & (0xff << 8 ), cond);
+    orr(dst, dst, imm32 & (0xff << 16), cond);
+    if(s) orrs(dst, dst, imm32 & (0xff << 24), cond);
+    else  orr (dst, dst, imm32 & (0xff << 24), cond);
   }
 }
 
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/assembler_aarch32.hpp
--- a/src/cpu/aarch32/vm/assembler_aarch32.hpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/assembler_aarch32.hpp	Fri Jan 15 23:23:56 2016 +0800
@@ -45,22 +45,22 @@
 class Argument VALUE_OBJ_CLASS_SPEC {
  public:
   enum {
-        n_int_register_parameters_c   = 4,  // r0, r1, ... r3 (c_rarg0, c_rarg1, ...)
-        // These are number of double registers, each double register can instead be
-        // used as two single precision registers.
-        #ifdef __VFP_FP__
-                n_float_register_parameters_c = 8,  // d0, d1, ... d7 (c_farg0, c_farg1, ... )
-                #define HARD_FLOAT_CC
-        #elif defined(__SOFTFP__)
-                n_float_register_parameters_c = 0,  // Not realy used, in this case
-                // pass_float -> pass_int & pass_double -> pass_long
-                #else
-                #error "Floating point ABI not supported";
-                #endif
+    n_int_register_parameters_c   = 4,  // r0, r1, ... r3 (c_rarg0, c_rarg1, ...)
+    // These are number of double registers, each double register can instead be
+    // used as two single precision registers.
+#ifdef __VFP_FP__
+    n_float_register_parameters_c = 8,  // d0, d1, ... d7 (c_farg0, c_farg1, ... )
+#define HARD_FLOAT_CC
+#elif defined(__SOFTFP__)
+    n_float_register_parameters_c = 0,  // Not realy used, in this case
+    // pass_float -> pass_int & pass_double -> pass_long
+#else
+  #error "Floating point ABI not supported";
+#endif
 
     // Not that these only make sense for compiled
-                // value are irrelevant currently, if c1, c2, ... will go ahead these need to be
-                // sorted out.
+    // value are irrelevant currently, if c1, c2, ... will go ahead these need to be
+    // sorted out.
     n_int_register_parameters_j   = 4, // r1, ... r7, r0 (rj_rarg0, j_rarg1, ...
     n_float_register_parameters_j = 4  // d0, d1, ... d7 (j_farg0, j_farg1, ...
   };
@@ -963,8 +963,8 @@
     wrap_label(Rd, L, cond, &Assembler::Assembler::adr);
   }
 
- private:
-        friend void entry(CodeBuffer *cb);
+private:
+  friend void entry(CodeBuffer *cb);
 #define INSN(NAME, decode, s_flg)                                                    \
   inline void NAME(Register Rd, unsigned imm, Condition cond = C_DFLT) {             \
     bool status = imm_instr(decode, Rd, ZERO_ADDR_REG, imm, cond, s_flg);            \
@@ -1599,24 +1599,24 @@
   // MCR<c> <coproc>, <opc1>, <Rt>, <CRn>, <CRm>{, <opc2>}
   void mcr(int cpc_dex, int opc1, Register Rt, int cpc_reg_dex1,
            int cpc_reg_dex2, int opc2, Condition cond = C_DFLT) {
-        starti;
-        f(cond, 31, 28), f(0b1110, 27, 24), f(opc1, 23, 21), f(0, 20);
-        f(cpc_reg_dex1, 19, 16), rf(Rt, 12), f(cpc_dex, 11, 8);
-        f(opc2, 7, 5), f(1, 4), f(cpc_reg_dex2, 3, 0);
+    starti;
+    f(cond, 31, 28), f(0b1110, 27, 24), f(opc1, 23, 21), f(0, 20);
+    f(cpc_reg_dex1, 19, 16), rf(Rt, 12), f(cpc_dex, 11, 8);
+    f(opc2, 7, 5), f(1, 4), f(cpc_reg_dex2, 3, 0);
   }
 
   // These instructions do not read the value of the register passed,
   // can be any. Chosen r0.
   void cp15dmb(Condition cond = C_DFLT) {
-        mcr(15, 0, r0, 7, 10, 5, cond);
+    mcr(15, 0, r0, 7, 10, 5, cond);
   }
 
   void cp15dsb(Condition cond = C_DFLT) {
-        mcr(15, 0, r0, 7, 10, 4, cond);
+    mcr(15, 0, r0, 7, 10, 4, cond);
   }
 
   void cp15isb(Condition cond = C_DFLT) {
-        mcr(15, 0, r0, 7, 5, 4, cond);
+    mcr(15, 0, r0, 7, 5, 4, cond);
   }
 
   enum Membar_mask_bits {
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/interp_masm_aarch32.cpp
--- a/src/cpu/aarch32/vm/interp_masm_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/interp_masm_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -422,48 +422,48 @@
     verify_oop(r0, state);
   }
 
-        /* Debugging code */
-        bytecode_seen(rscratch1, r3);
+  /* Debugging code */
+  bytecode_seen(rscratch1, r3);
 
-        /*{
-                Label skip;
+  /*{
+    Label skip;
 
-                mov(r3, (address)&MacroAssembler::bytecodes_executed);
-                ldr(r2, r3);
-                add(r2, r2, 1);
-                str(r2, r3);
-                // Print out every 16384 (needs to be a power of two).
-                mov(r3, 16384 - 1);
-                tst(r2, r3);
-                b(skip, Assembler::NE);
-                reg_printf_important("Executed %d bytecodes.\n", r2);
-                bind(skip);
-        }*/
+    mov(r3, (address)&MacroAssembler::bytecodes_executed);
+    ldr(r2, r3);
+    add(r2, r2, 1);
+    str(r2, r3);
+    // Print out every 16384 (needs to be a power of two).
+    mov(r3, 16384 - 1);
+    tst(r2, r3);
+    b(skip, Assembler::NE);
+    reg_printf_important("Executed %d bytecodes.\n", r2);
+    bind(skip);
+  }*/
 
 
-        /*mov(r3, (address)&MacroAssembler::bytecodes_until_print);
-        ldr(r2, Address(r3));
-        cmp(r2, 0);
+  /*mov(r3, (address)&MacroAssembler::bytecodes_until_print);
+  ldr(r2, Address(r3));
+  cmp(r2, 0);
 
-        sub(r2, r2, 1, Assembler::NE);
-        str(r2, Address(r3), Assembler::NE);
+  sub(r2, r2, 1, Assembler::NE);
+  str(r2, Address(r3), Assembler::NE);
 
-        mov(r2, 1, Assembler::EQ);
-        mov(r3, (address)&MacroAssembler::enable_debug, Assembler::EQ);
-        str(r2, Address(r3), Assembler::EQ);
+  mov(r2, 1, Assembler::EQ);
+  mov(r3, (address)&MacroAssembler::enable_debug, Assembler::EQ);
+  str(r2, Address(r3), Assembler::EQ);
 
-        mov(r3, (address)&MacroAssembler::enable_method_debug, Assembler::EQ);
-        str(r2, Address(r3), Assembler::EQ);*/
+  mov(r3, (address)&MacroAssembler::enable_method_debug, Assembler::EQ);
+  str(r2, Address(r3), Assembler::EQ);*/
 
-        /*Label end;
-        cmp(r2, 0);
-        b(end, Assembler::NE);
-        stop("got to end of bytecodes");
-        bind(end);*/
+  /*Label end;
+  cmp(r2, 0);
+  b(end, Assembler::NE);
+  stop("got to end of bytecodes");
+  bind(end);*/
 
-        get_bytecode(r14, rscratch1);
-        reg_printf("Dispatching bytecode %s (%d) @ BCP = %p\n", r14, rscratch1, rbcp);
-        /* End debugging code */
+  get_bytecode(r14, rscratch1);
+  reg_printf("Dispatching bytecode %s (%d) @ BCP = %p\n", r14, rscratch1, rbcp);
+  /* End debugging code */
 
 
   if (table == Interpreter::dispatch_table(state)) {
@@ -580,7 +580,7 @@
   bind(unlocked);
 
   // r0: Might contain return value
-        // FIXME r1 : Might contain the value too
+  // FIXME r1 : Might contain the value too
 
   // Check that all monitors are unlocked
   {
@@ -788,7 +788,7 @@
   } else {
     Label done;
 
-                //create_breakpoint();
+    //create_breakpoint();
     const Register swap_reg   = c_rarg0;
     const Register header_reg = c_rarg2;  // Will contain the old oopMark
     const Register obj_reg    = c_rarg3;  // Will contain the oop
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/interpreterRT_aarch32.cpp
--- a/src/cpu/aarch32/vm/interpreterRT_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/interpreterRT_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -39,9 +39,9 @@
 #define __ _masm->
 
 /*#define print_copy(name, off) \
-        __ mov(rscratch1, (address)name);\
-        __ mov(rscratch2, off);\
-        __ reg_printf("%s copied from offset %p + %d\n", rscratch1, from(), rscratch2);*/
+  __ mov(rscratch1, (address)name);\
+  __ mov(rscratch2, off);\
+  __ reg_printf("%s copied from offset %p + %d\n", rscratch1, from(), rscratch2);*/
 
 #define print_copy(name, off)
 
@@ -51,7 +51,7 @@
 Register InterpreterRuntime::SignatureHandlerGenerator::temp() { return rscratch1; }
 
 void InterpreterRuntime::SignatureHandlerGenerator::pass_int() {
-        print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
+  print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
   const Address src(from(), Interpreter::local_offset_in_bytes(offset()));
 
   switch (_num_int_args) {
@@ -77,7 +77,7 @@
 }
 
 void InterpreterRuntime::SignatureHandlerGenerator::pass_long() {
-        print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset() + 1));
+  print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset() + 1));
   const Address src(from(), Interpreter::local_offset_in_bytes(offset() + 1));
   // Needs to be aligned to even registers. Means also won't be split across
   // registers and stack.
@@ -90,7 +90,7 @@
     break;
   default:
     __ ldrd(r0, temp(), src);
-        _stack_offset = (_stack_offset + 7) & ~7; // Align on 8-byte boundary
+    _stack_offset = (_stack_offset + 7) & ~7; // Align on 8-byte boundary
     __ strd(r0, temp(), Address(to(), _stack_offset));
     _stack_offset += 2 * wordSize;
     _num_int_args += 2;
@@ -100,13 +100,13 @@
 
 #ifdef HARD_FLOAT_CC
 void InterpreterRuntime::SignatureHandlerGenerator::pass_float() {
-        print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
+  print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
   const Address src(from(), Interpreter::local_offset_in_bytes(offset()));
 
   if (_fp_arg_mask) {
-        unsigned index = __builtin_ctz(_fp_arg_mask);
-        __ vldr_f32_real(index, src);
-        _fp_arg_mask ^= 1 << index;
+    unsigned index = __builtin_ctz(_fp_arg_mask);
+    __ vldr_f32_real(index, src);
+    _fp_arg_mask ^= 1 << index;
     _next_double_dex += index & 1;
   } else {
     __ ldr(r0, src);
@@ -116,33 +116,33 @@
 }
 
 void InterpreterRuntime::SignatureHandlerGenerator::pass_double() {
-        print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset() + 1));
+  print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset() + 1));
   const Address src(from(), Interpreter::local_offset_in_bytes(offset() + 1));
 
   if (_next_double_dex < Argument::n_float_register_parameters_c) {
-                _fp_arg_mask ^= 3 << _next_double_dex * 2;
-                __ vldr_f64(as_FloatRegister(_next_double_dex++), src);
+    _fp_arg_mask ^= 3 << _next_double_dex * 2;
+    __ vldr_f64(as_FloatRegister(_next_double_dex++), src);
   } else {
-                __ ldrd(r0, temp(), src);
-                _stack_offset = (_stack_offset + 7) & ~7;
-                __ strd(r0, temp(), Address(to(), _stack_offset));
-                _stack_offset += 2 * wordSize;
-        }
+    __ ldrd(r0, temp(), src);
+    _stack_offset = (_stack_offset + 7) & ~7;
+    __ strd(r0, temp(), Address(to(), _stack_offset));
+    _stack_offset += 2 * wordSize;
+  }
 }
 #else
 // Just pass them in integer registers and on the stack as we would
 // any other argument
 void InterpreterRuntime::SignatureHandlerGenerator::pass_float() {
-        pass_int();
+  pass_int();
 }
 
 void InterpreterRuntime::SignatureHandlerGenerator::pass_double() {
-        pass_long();
+  pass_long();
 }
 #endif //HARD_FLOAT_CC
 
 void InterpreterRuntime::SignatureHandlerGenerator::pass_object() {
-        print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
+  print_copy(__FUNCTION__, Interpreter::local_offset_in_bytes(offset()));
 
   switch (_num_int_args) {
   case 0:
@@ -234,22 +234,22 @@
 
   virtual void pass_long()
   {
-                intptr_t high_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(0));
+    intptr_t high_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(0));
     intptr_t low_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(1));
     _from -= 2*Interpreter::stackElementSize;
 
     if (_num_int_reg_args < Argument::n_int_register_parameters_c-2) {
-        // Passing longs. As c_rarg0 is always reserved for jni_env we could only
-        // possibly stash a long in r3:r2 due to alignment so we can only enter here
-        // with either zero or one parameters.
-        // Align to two
-        _int_args += 1 - _num_int_reg_args; // 0 or 1
+      // Passing longs. As c_rarg0 is always reserved for jni_env we could only
+      // possibly stash a long in r3:r2 due to alignment so we can only enter here
+      // with either zero or one parameters.
+      // Align to two
+      _int_args += 1 - _num_int_reg_args; // 0 or 1
       *_int_args++ = low_obj;
       *_int_args++ = high_obj;
       _num_int_reg_args = 3;
     } else {
-            _to = (intptr_t*)(((intptr_t)_to + 7) & ~7); // Align to eight bytes
-        *_to++ = low_obj;
+      _to = (intptr_t*)(((intptr_t)_to + 7) & ~7); // Align to eight bytes
+      *_to++ = low_obj;
       *_to++ = high_obj;
       _num_int_reg_args = 3;
     }
@@ -274,7 +274,7 @@
     _from -= Interpreter::stackElementSize;
 
     if (_fp_arg_mask) {
-        unsigned index = __builtin_ctz(_fp_arg_mask);
+      unsigned index = __builtin_ctz(_fp_arg_mask);
       _fp_args[index] = from_obj;
       _fp_arg_mask ^= 1 << index;
       _next_double_dex += index & 1;
@@ -285,7 +285,7 @@
 
   virtual void pass_double()
   {
-                intptr_t high_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(0));
+    intptr_t high_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(0));
     intptr_t low_obj = *(intptr_t*)(_from+Interpreter::local_offset_in_bytes(1));
     _from -= 2*Interpreter::stackElementSize;
 
@@ -293,13 +293,13 @@
       //We can allocate to a register.
       int index = 2 * (_next_double_dex++);
       _fp_arg_mask ^= 3 << index;
-            _fp_args[index] = low_obj;
-            _fp_args[index + 1] = high_obj;
+      _fp_args[index] = low_obj;
+      _fp_args[index + 1] = high_obj;
     } else {
-                  _to = (intptr_t*)(((intptr_t)_to + 7) & ~7); // Align to eight bytes
-                        *_to++ = low_obj;
-                  *_to++ = high_obj;
-                }
+      _to = (intptr_t*)(((intptr_t)_to + 7) & ~7); // Align to eight bytes
+      *_to++ = low_obj;
+      *_to++ = high_obj;
+    }
   }
 #else
   virtual void pass_float() { pass_int(); }
@@ -312,7 +312,7 @@
   {
     _from = from;
     _to   = to;
-                // See layout in interpreter_aarch32.cpp
+    // See layout in interpreter_aarch32.cpp
     _int_args = to - (method->is_static() ? 19 : 20);
     _fp_args =  to - 16; //each slot is for a double
     _fp_identifiers = to - 21;
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/interpreter_aarch32.cpp
--- a/src/cpu/aarch32/vm/interpreter_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/interpreter_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -58,8 +58,8 @@
 address AbstractInterpreterGenerator::generate_slow_signature_handler() {
   address entry = __ pc();
 
-        // The sp should be aligned on entry to the bottom of where the integer args
-        // need to be copied to.
+  // The sp should be aligned on entry to the bottom of where the integer args
+  // need to be copied to.
 
   // rmethod
   // rlocals
@@ -128,7 +128,7 @@
   // r13: sender sp
   // esp: args
 
-        //if (!InlineIntrinsics) return NULL; // Generate a vanilla entry
+  //if (!InlineIntrinsics) return NULL; // Generate a vanilla entry
   // FIXME currently ignoring this flag and inlining anyway
 
   // These don't need a safepoint check because they aren't virtually
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/jniFastGetField_aarch32.cpp
--- a/src/cpu/aarch32/vm/jniFastGetField_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/jniFastGetField_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -49,8 +49,8 @@
   Register result   = c_rarg0;
   Register robj     = c_rarg1;
   Register rcounter = c_rarg3;
-        int      args     = RegSet::of(c_rarg0, c_rarg1, c_rarg2).bits();
-        int      nargs    = 3;
+  int      args     = RegSet::of(c_rarg0, c_rarg1, c_rarg2).bits();
+  int      nargs    = 3;
 
   const char *name;
   switch (type) {
@@ -75,8 +75,8 @@
 
   Label slow;
 
-        __ lea(rcounter, SafepointSynchronize::safepoint_counter_addr());
-        __ ldr(rcounter, rcounter);
+  __ lea(rcounter, SafepointSynchronize::safepoint_counter_addr());
+  __ ldr(rcounter, rcounter);
   __ tst(rcounter, 1);
   __ b(slow, Assembler::NE);
   __ stmdb(sp, args);
@@ -95,16 +95,16 @@
     case T_INT:     __ ldr (result, Address(robj, c_rarg2, lsr(2))); break;
     case T_BOOLEAN: __ ldrb(result, Address(robj, c_rarg2, lsr(2))); break;
     default: {
-        __ lsr(c_rarg2, c_rarg2, 2);
-        switch(type) {
-                                case T_BYTE:    __ ldrsb   (result, Address(robj, c_rarg2)); break;
-                                case T_CHAR:    __ ldrh    (result, Address(robj, c_rarg2)); break;
-                                case T_SHORT:   __ ldrsh   (result, Address(robj, c_rarg2)); break;
-                                case T_DOUBLE:
-                                case T_LONG:    __ ldrd    (result, Address(robj, c_rarg2)); break;
-                                default:        ShouldNotReachHere();
-                        }
-                }
+      __ lsr(c_rarg2, c_rarg2, 2);
+      switch(type) {
+        case T_BYTE:    __ ldrsb   (result, Address(robj, c_rarg2)); break;
+        case T_CHAR:    __ ldrh    (result, Address(robj, c_rarg2)); break;
+        case T_SHORT:   __ ldrsh   (result, Address(robj, c_rarg2)); break;
+        case T_DOUBLE:
+        case T_LONG:    __ ldrd    (result, Address(robj, c_rarg2)); break;
+        default:        ShouldNotReachHere();
+      }
+    }
   }
   __ lea(rscratch2, SafepointSynchronize::safepoint_counter_addr());
   // rscratch2 is address dependent on result.
@@ -123,7 +123,7 @@
   __ b(lr, Assembler::EQ);
 
   // Restore args for slowcase call into the vm
-        __ ldmia(sp, args);
+  __ ldmia(sp, args);
 
   // Slowcase
   slowcase_entry_pclist[count++] = __ pc();
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/jniTypes_aarch32.hpp
--- a/src/cpu/aarch32/vm/jniTypes_aarch32.hpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/jniTypes_aarch32.hpp	Fri Jan 15 23:23:56 2016 +0800
@@ -65,26 +65,26 @@
     *(jlong*) (to + 1 + pos) = *from;
     pos += 2;
   }*/
-        static inline void put_long(jlong  from, intptr_t *to) {
-                uint64_t val = from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+  static inline void put_long(jlong  from, intptr_t *to) {
+    uint64_t val = from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
     *(jlong*)to = (jlong)val;
   }
 
   static inline void put_long(jlong  from, intptr_t *to, int& pos) {
-                uint64_t val = from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+    uint64_t val = from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
 
     *(jlong*) (to + pos) = (jlong)val;
     pos += 2;
   }
 
   static inline void put_long(jlong *from, intptr_t *to, int& pos) {
-                uint64_t val = *from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+    uint64_t val = *from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
 
     *(jlong*) (to + pos) = (jlong)val;
     pos += 2;
@@ -121,24 +121,24 @@
   }*/
 
   static inline void put_double(jdouble  from, intptr_t *to) {
-                uint64_t val = *(uint64_t*)&from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+    uint64_t val = *(uint64_t*)&from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
     *(uint64_t*)to = val;
   }
 
   static inline void put_double(jdouble  from, intptr_t *to, int& pos) {
-                uint64_t val = *(uint64_t*)&from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+    uint64_t val = *(uint64_t*)&from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
     *(uint64_t*) (to + pos) = val;
     pos += 2;
   }
 
   static inline void put_double(jdouble *from, intptr_t *to, int& pos) {
-                uint64_t val = *(uint64_t*)from;
-                uint64_t mask = (1LL << 32) - 1;
-                val = (val >> 32) | ((val & mask) << 32);
+    uint64_t val = *(uint64_t*)from;
+    uint64_t mask = (1LL << 32) - 1;
+    val = (val >> 32) | ((val & mask) << 32);
     *(uint64_t*) (to + pos) = val;
     pos += 2;
   }
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/macroAssembler_aarch32.cpp
--- a/src/cpu/aarch32/vm/macroAssembler_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/macroAssembler_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -85,7 +85,7 @@
   } else if (0b0011 == opc) {
     // Movw, Movt or mov, orr, orr, orr
     // patch up address load to registers (absolute address).
-          instructions = patch_oop(branch, target) / NativeInstruction::instruction_size;
+    instructions = patch_oop(branch, target) / NativeInstruction::instruction_size;
   } else if (0b010 == (opc >> 1)) {
     // LDR, LDRB, STR, STRB
     Instruction_aarch32::patch(branch, 11, 0, uabs(offset));
@@ -114,34 +114,34 @@
 }
 
 int MacroAssembler::patch_oop(address insn_addr, address o) {
-        unsigned insn = *(unsigned*)insn_addr;
-        int opc = Instruction_aarch32::extract(insn, 27, 21);
-        if(0b0011000 == opc) {
-                //32-bit pointers, formed of a mov and a movt
-                assert(nativeInstruction_at(insn_addr+4)->is_movt(), "wrong insns in patch");
-
-                uint32_t btm = (uint32_t)o & 0xffff;
-                Instruction_aarch32::patch(insn_addr, 19, 16, btm >> 12);
-                Instruction_aarch32::patch(insn_addr, 11, 0, btm & 0xfff);
-                uint32_t top = (uint32_t)o >> 16;
-                Instruction_aarch32::patch(insn_addr + 4, 19, 16, top >> 12);
-                Instruction_aarch32::patch(insn_addr + 4, 11, 0, top & 0xfff);
-                return 2 * NativeInstruction::instruction_size;
+  unsigned insn = *(unsigned*)insn_addr;
+  int opc = Instruction_aarch32::extract(insn, 27, 21);
+  if(0b0011000 == opc) {
+    //32-bit pointers, formed of a mov and a movt
+    assert(nativeInstruction_at(insn_addr+4)->is_movt(), "wrong insns in patch");
+
+    uint32_t btm = (uint32_t)o & 0xffff;
+    Instruction_aarch32::patch(insn_addr, 19, 16, btm >> 12);
+    Instruction_aarch32::patch(insn_addr, 11, 0, btm & 0xfff);
+    uint32_t top = (uint32_t)o >> 16;
+    Instruction_aarch32::patch(insn_addr + 4, 19, 16, top >> 12);
+    Instruction_aarch32::patch(insn_addr + 4, 11, 0, top & 0xfff);
+    return 2 * NativeInstruction::instruction_size;
   } else if(0b0011101 == opc) {
-        //Instead 32bit load sequence uses mov, orr, orr, orr
-        assert(nativeInstruction_at(insn_addr+4 )->is_orr(), "wrong insns in patch");
-        assert(nativeInstruction_at(insn_addr+8 )->is_orr(), "wrong insns in patch");
-        assert(nativeInstruction_at(insn_addr+12)->is_orr(), "wrong insns in patch");
-        // FIXME this could carry us outside valid memory
-
-        uint32_t addr = (uint32_t)o;
-        Instruction_aarch32::patch(insn_addr + 0,  11, 0, (0b0000 << 8) | ((addr >>  0) & 0xff));
-        Instruction_aarch32::patch(insn_addr + 4,  11, 0, (0b1100 << 8) | ((addr >>  8) & 0xff));
-        Instruction_aarch32::patch(insn_addr + 8,  11, 0, (0b1000 << 8) | ((addr >> 16) & 0xff));
-        Instruction_aarch32::patch(insn_addr + 12, 11, 0, (0b0100 << 8) | ((addr >> 24) & 0xff));
-        return 4 * NativeInstruction::instruction_size;
+    //Instead 32bit load sequence uses mov, orr, orr, orr
+    assert(nativeInstruction_at(insn_addr+4 )->is_orr(), "wrong insns in patch");
+    assert(nativeInstruction_at(insn_addr+8 )->is_orr(), "wrong insns in patch");
+    assert(nativeInstruction_at(insn_addr+12)->is_orr(), "wrong insns in patch");
+    // FIXME this could carry us outside valid memory
+
+    uint32_t addr = (uint32_t)o;
+    Instruction_aarch32::patch(insn_addr + 0,  11, 0, (0b0000 << 8) | ((addr >>  0) & 0xff));
+    Instruction_aarch32::patch(insn_addr + 4,  11, 0, (0b1100 << 8) | ((addr >>  8) & 0xff));
+    Instruction_aarch32::patch(insn_addr + 8,  11, 0, (0b1000 << 8) | ((addr >> 16) & 0xff));
+    Instruction_aarch32::patch(insn_addr + 12, 11, 0, (0b0100 << 8) | ((addr >> 24) & 0xff));
+    return 4 * NativeInstruction::instruction_size;
   } else {
-        ShouldNotReachHere();
+    ShouldNotReachHere();
   }
   return 0; //won't reach here
 }
@@ -157,28 +157,28 @@
     unsigned *insn_buf = (unsigned*)insn_addr;
     int opc2 = Instruction_aarch32::extract(insn, 23, 21);
     if(0b000 == opc2) {
-        // movw, movt (only on newer ARMs)
-                  assert(nativeInstruction_at(&insn_buf[1])->is_movt(), "wrong insns in patch");
-                  u_int32_t addr;
-                  addr  = Instruction_aarch32::extract(insn_buf[1], 19, 16) << 28;
-                  addr |= Instruction_aarch32::extract(insn_buf[1], 11, 0) << 16;
-                  addr |= Instruction_aarch32::extract(insn_buf[0], 19, 16) << 12;
-                  addr |= Instruction_aarch32::extract(insn_buf[0], 11, 0);
-                  return address(addr);
-                } else if(0b101 == opc2) {
-                  // mov, orr, orr, orr
-                        assert(nativeInstruction_at(&insn_buf[1])->is_orr(), "wrong insns in patch");
-                        assert(nativeInstruction_at(&insn_buf[2])->is_orr(), "wrong insns in patch");
-                        assert(nativeInstruction_at(&insn_buf[3])->is_orr(), "wrong insns in patch");
-                        u_int32_t addr;
-                  // TODO Check that the rotations are in the expected order.
-                        addr  = Instruction_aarch32::extract(insn_buf[0], 7, 0) << 0;
-                  addr |= Instruction_aarch32::extract(insn_buf[1], 7, 0) << 8;
-                  addr |= Instruction_aarch32::extract(insn_buf[2], 7, 0) << 16;
-                  addr |= Instruction_aarch32::extract(insn_buf[3], 7, 0) << 24;
-                  return address(addr);
+      // movw, movt (only on newer ARMs)
+      assert(nativeInstruction_at(&insn_buf[1])->is_movt(), "wrong insns in patch");
+      u_int32_t addr;
+      addr  = Instruction_aarch32::extract(insn_buf[1], 19, 16) << 28;
+      addr |= Instruction_aarch32::extract(insn_buf[1], 11, 0) << 16;
+      addr |= Instruction_aarch32::extract(insn_buf[0], 19, 16) << 12;
+      addr |= Instruction_aarch32::extract(insn_buf[0], 11, 0);
+      return address(addr);
+    } else if(0b101 == opc2) {
+      // mov, orr, orr, orr
+      assert(nativeInstruction_at(&insn_buf[1])->is_orr(), "wrong insns in patch");
+      assert(nativeInstruction_at(&insn_buf[2])->is_orr(), "wrong insns in patch");
+      assert(nativeInstruction_at(&insn_buf[3])->is_orr(), "wrong insns in patch");
+      u_int32_t addr;
+      // TODO Check that the rotations are in the expected order.
+      addr  = Instruction_aarch32::extract(insn_buf[0], 7, 0) << 0;
+      addr |= Instruction_aarch32::extract(insn_buf[1], 7, 0) << 8;
+      addr |= Instruction_aarch32::extract(insn_buf[2], 7, 0) << 16;
+      addr |= Instruction_aarch32::extract(insn_buf[3], 7, 0) << 24;
+      return address(addr);
     } else {
-        ShouldNotReachHere();
+      ShouldNotReachHere();
     }
   } else if (0b010 == (opc >> 1)) {
     // LDR, LDRB, STR, STRB
@@ -570,8 +570,8 @@
   set_last_Java_frame(last_java_sp, rfp, l, rscratch2);
 
 
-        // FIXME - Can save lr in more elegant way ?
-        //str(lr, pre(sp, -wordSize));
+  // FIXME - Can save lr in more elegant way ?
+  //str(lr, pre(sp, -wordSize));
 
   // do the call, remove parameters
   MacroAssembler::call_VM_leaf_base(entry_point, number_of_arguments, &l);
@@ -1591,7 +1591,7 @@
 
 void MacroAssembler::debug32(char* msg, int32_t pc, int32_t regs[])
 {
-        print_unseen_bytecodes();
+  print_unseen_bytecodes();
   // In order to get locks to work, we need to fake a in_VM state
   if (ShowMessageBoxOnError) {
     JavaThread* thread = JavaThread::current();
@@ -2238,57 +2238,57 @@
 }
 
 void MacroAssembler::divide32(Register res, Register num, Register den, bool want_mod) {
-        Register cnt = rscratch1;
-        Register mod = rscratch2;
-        Register sign = r14;
-        assert_different_registers(num, den, rscratch1, rscratch2, r14);
-
-        // FIXME This works by first converting any negative values to positive ones, however
-        // it is not possible to express |INT_MIN|. Need to fix this
-
-        //Convert to positive values
-        mov(sign, 0);
-
-        cmp(num, 0);
-        mov(sign, 1, MI);
-        rsb(num, num, 0, MI);
-
-        cmp(den, 0);
-        if(!want_mod) eor(sign, sign, 1, MI);
-        rsb(den, den, 0, MI);
-
-        // Algorithm from
-        // http://www.chiark.greenend.org.uk/~theom/riscos/docs/ultimate/a252div.txt
-        // Graeme Williams
-        mov(cnt, 28);
-        mov(mod, num, lsr(4));
-        cmp(den, mod, lsr(12));
-        sub(cnt, cnt, 16, Assembler::LE);
-        mov(mod, mod, lsr(16), Assembler::LE);
-        cmp(den, mod, lsr(4));
-        sub(cnt, cnt, 8, Assembler::LE);
-        mov(mod, mod, lsr(8), Assembler::LE);
-        cmp(den, mod);
-        sub(cnt, cnt, 4, Assembler::LE);
-        mov(mod, mod, lsr(4), Assembler::LE);
-        mov(num, num, lsl(cnt));
-        rsb(den, den, 0);
-
-        adds(num, num, num);
-        //Now skip over cnt copies of the 3 instr. loop.
-        add(cnt, cnt, cnt, lsl(1));
-        add(r15_pc, r15_pc, cnt, lsl(2));
-        mov(r0, r0);
-
-        for(int i = 0; i < 32; i++) {
-                adcs(mod, den, mod, lsl(1));
-                sub(mod, mod, den, Assembler::LO);
-                adcs(num, num, num);
-        }
-
-        cmp(sign, 0);
-        rsb(res, want_mod? mod : num, 0, NE);
-        mov(res, want_mod? mod : num, EQ);
+  Register cnt = rscratch1;
+  Register mod = rscratch2;
+  Register sign = r14;
+  assert_different_registers(num, den, rscratch1, rscratch2, r14);
+
+  // FIXME This works by first converting any negative values to positive ones, however
+  // it is not possible to express |INT_MIN|. Need to fix this
+
+  //Convert to positive values
+  mov(sign, 0);
+
+  cmp(num, 0);
+  mov(sign, 1, MI);
+  rsb(num, num, 0, MI);
+
+  cmp(den, 0);
+  if(!want_mod) eor(sign, sign, 1, MI);
+  rsb(den, den, 0, MI);
+
+  // Algorithm from
+  // http://www.chiark.greenend.org.uk/~theom/riscos/docs/ultimate/a252div.txt
+  // Graeme Williams
+  mov(cnt, 28);
+  mov(mod, num, lsr(4));
+  cmp(den, mod, lsr(12));
+  sub(cnt, cnt, 16, Assembler::LE);
+  mov(mod, mod, lsr(16), Assembler::LE);
+  cmp(den, mod, lsr(4));
+  sub(cnt, cnt, 8, Assembler::LE);
+  mov(mod, mod, lsr(8), Assembler::LE);
+  cmp(den, mod);
+  sub(cnt, cnt, 4, Assembler::LE);
+  mov(mod, mod, lsr(4), Assembler::LE);
+  mov(num, num, lsl(cnt));
+  rsb(den, den, 0);
+
+  adds(num, num, num);
+  //Now skip over cnt copies of the 3 instr. loop.
+  add(cnt, cnt, cnt, lsl(1));
+  add(r15_pc, r15_pc, cnt, lsl(2));
+  mov(r0, r0);
+
+  for(int i = 0; i < 32; i++) {
+    adcs(mod, den, mod, lsl(1));
+    sub(mod, mod, den, Assembler::LO);
+    adcs(num, num, num);
+  }
+
+  cmp(sign, 0);
+  rsb(res, want_mod? mod : num, 0, NE);
+  mov(res, want_mod? mod : num, EQ);
 }
 
 
@@ -2297,7 +2297,7 @@
 // <Rd> = <Rn> / <Rm>
 // <Rd> = <Rn> % <Rm>
 void MacroAssembler::divide(Register Rd, Register Rn, Register Rm, int width, bool want_remainder) {
-        //Dispatch to best possible
+  //Dispatch to best possible
   Register Rdh = (Register)(Rd->encoding_nocheck() + 1);
   Register Rnh = (Register)(Rn->encoding_nocheck() + 1);
   Register Rmh = (Register)(Rm->encoding_nocheck() + 1);
@@ -2306,75 +2306,75 @@
   bool is64b = 64 == width;
 
   if(is64b) {
-        assert_different_registers(Rn, Rnh, Rm, Rmh, rscratch1, rscratch2);
+    assert_different_registers(Rn, Rnh, Rm, Rmh, rscratch1, rscratch2);
   } else {
-        assert_different_registers(Rn, Rm, rscratch1, rscratch2, r14); // r14 used by divide32
+    assert_different_registers(Rn, Rm, rscratch1, rscratch2, r14); // r14 used by divide32
   }
 
   if(!is64b && VM_Version::features() & FT_HW_DIVIDE) {
-        // Emit a hw instruction sequnce.
-        if(want_remainder) {
-                sdiv(rscratch1, Rn, Rm);
-                mls(Rd, rscratch1, Rm, Rn);
-        } else {
-                sdiv(Rd, Rn, Rm);
-        }
+    // Emit a hw instruction sequnce.
+    if(want_remainder) {
+      sdiv(rscratch1, Rn, Rm);
+      mls(Rd, rscratch1, Rm, Rn);
+    } else {
+      sdiv(Rd, Rn, Rm);
+    }
   } else if(!is64b) {
-        // Fall back to assembly software routine
-        divide32(Rd, Rn, Rm, want_remainder);
+    // Fall back to assembly software routine
+    divide32(Rd, Rn, Rm, want_remainder);
   } else {
-        // Fall back to C software routine for
-        // 64 bit divide/mod
-                if(Rn != r0) {
-                        mov(rscratch1, Rm);
-                        mov(rscratch2, Rmh);
-
-                        mov(r0, Rn);
-                        mov(r1, Rnh);
-
-                        mov(r2, rscratch1);
-                        mov(r3, rscratch2);
-                } else if(Rm != r2) {
-                        mov(r2, Rm);
-                        mov(r3, Rmh);
-                }
-                address function;
-                if(want_remainder) function = (address)internal_lmod;
-                else               function = (address)internal_ldiv;
-
-                mov(rscratch1, function);
-                bl(rscratch1);
-                if(Rd != r0) {
-                  mov(Rd, r0);
-                  if(is64b) mov(Rdh, r1);
-                }
+    // Fall back to C software routine for
+    // 64 bit divide/mod
+    if(Rn != r0) {
+      mov(rscratch1, Rm);
+      mov(rscratch2, Rmh);
+
+      mov(r0, Rn);
+      mov(r1, Rnh);
+
+      mov(r2, rscratch1);
+      mov(r3, rscratch2);
+    } else if(Rm != r2) {
+      mov(r2, Rm);
+      mov(r3, Rmh);
+    }
+    address function;
+    if(want_remainder) function = (address)internal_lmod;
+    else               function = (address)internal_ldiv;
+
+    mov(rscratch1, function);
+    bl(rscratch1);
+    if(Rd != r0) {
+      mov(Rd, r0);
+      if(is64b) mov(Rdh, r1);
+    }
   }
 }
 
 void MacroAssembler::extract_bits(Register dest, Register source, int lsb, int width) {
-        assert(lsb >= 0 && lsb + width <= 32 && width != 0, "Invalid lsb/width");
-        // Dispatch to the best sequence
-        if(0 == (lsb & 7) && (width == 8 || width == 16 || width == 32)) {
-                // Can use extend X
-          switch(width){
-                case 8:  uxtb(dest, source, ror(lsb)); break;
-                case 16: uxth(dest, source, ror(lsb)); break;
-                default:                               break;
-         }
-        } else if(VM_Version::features() & (FT_ARMV7 | FT_ARMV6T2)) {
-          ubfx(dest, source, lsb, width);
-        } else {
-                // Do two shifts
-                lsl(dest, source, 32 - (width + lsb));
-                lsr(dest, dest, 32 - width);
-        }
+  assert(lsb >= 0 && lsb + width <= 32 && width != 0, "Invalid lsb/width");
+  // Dispatch to the best sequence
+  if(0 == (lsb & 7) && (width == 8 || width == 16 || width == 32)) {
+    // Can use extend X
+    switch(width){
+      case 8:  uxtb(dest, source, ror(lsb)); break;
+      case 16: uxth(dest, source, ror(lsb)); break;
+      default:                               break;
+   }
+  } else if(VM_Version::features() & (FT_ARMV7 | FT_ARMV6T2)) {
+    ubfx(dest, source, lsb, width);
+  } else {
+    // Do two shifts
+    lsl(dest, source, 32 - (width + lsb));
+    lsr(dest, dest, 32 - width);
+  }
 }
 
 
 void MacroAssembler::atomic_ldrd(Register Rt, Register Rt2, Register Rbase) {
   assert(Rt->encoding_nocheck() % 2 == 0, "Must be an even register");
   assert((Register) (Rt + 1) == Rt2, "Must be contiguous");
-  if (VM_Version::features() & FT_SINGLE_CORE) {
+  if(VM_Version::features() & FT_SINGLE_CORE) {
     ldrd(Rt, Rbase);
   } else if (VM_Version::features() & (FT_ARMV7 | FT_ARMV6K)) {
 #ifdef ASSERT
@@ -2389,7 +2389,7 @@
   } else {
     // TODO: Find Java way of logging
     static bool warning_printed = false;
-    if (!warning_printed) {
+    if(!warning_printed) {
       fprintf(stderr, "Unable to provide atomic doubleword load.\n");
       warning_printed = true;
     }
@@ -2403,7 +2403,7 @@
   assert((Register) (Rt + 1) == Rt2, "Must be contiguous");
   assert((Register) (temp + 1) == temp2, "Must be contiguous");
   assert_different_registers(temp, Rt, Rbase, temp2);
-  if (VM_Version::features() & FT_SINGLE_CORE) {
+  if(VM_Version::features() & FT_SINGLE_CORE) {
     strd(Rt, Rbase);
   } else if (VM_Version::features() & (FT_ARMV7 | FT_ARMV6K)) {
     // First need to gain exclusive access
@@ -2423,7 +2423,7 @@
   } else {
     // TODO: Find Java way of logging
     static bool warning_printed = false;
-    if (!warning_printed) {
+    if(!warning_printed) {
       fprintf(stderr, "Unable to provide atomic doubleword store.\n");
       warning_printed = true;
     }
@@ -2470,49 +2470,49 @@
 int bytecodes_seen[256];
 
 void MacroAssembler::init_unseen_bytecodes() {
-        for(int i = 0; i < 256; i++ ) {
-                bytecodes_seen[i] = 0;
-        }
+  for(int i = 0; i < 256; i++ ) {
+    bytecodes_seen[i] = 0;
+  }
 }
 
 void MacroAssembler::bytecode_seen(Register bc_reg, Register scratch) {
-        if(ENABLE_DEBUGGING) {
-                mov(scratch, (address)bytecodes_seen);
-                add(scratch, scratch, bc_reg, lsl(2));
-                add(bc_reg, bc_reg, 1);
-                str(bc_reg, Address(scratch));
-                sub(bc_reg, bc_reg, 1);
-        }
+  if(ENABLE_DEBUGGING) {
+    mov(scratch, (address)bytecodes_seen);
+    add(scratch, scratch, bc_reg, lsl(2));
+    add(bc_reg, bc_reg, 1);
+    str(bc_reg, Address(scratch));
+    sub(bc_reg, bc_reg, 1);
+  }
 }
 
 void MacroAssembler::print_unseen_bytecodes() {
-        if(ENABLE_DEBUGGING) {
-                printf("=== Unseen bytecodes ===\n");
-                for(int i = 0; i < N_J_BYTECODES; i++) {
-                        if(0 == bytecodes_seen[i]) {
-                                printf("\t%s\n", j_bytecodes[i]);
-                        }
-                }
-                printf("=== End unseen ===\n");
-        } else {
-                printf("Not kept track, enable debugging to view info\n");
-        }
-        fflush(stdout);
+  if(ENABLE_DEBUGGING) {
+    printf("=== Unseen bytecodes ===\n");
+    for(int i = 0; i < N_J_BYTECODES; i++) {
+      if(0 == bytecodes_seen[i]) {
+        printf("\t%s\n", j_bytecodes[i]);
+      }
+    }
+    printf("=== End unseen ===\n");
+  } else {
+    printf("Not kept track, enable debugging to view info\n");
+  }
+  fflush(stdout);
 }
 
 int machine_state_regset = 0b0101111111111111;
 int machine_state_float_regset = 0b11;
 
 void MacroAssembler::save_machine_state() {
-        stmdb(sp, machine_state_regset);
-        vstmdb_f64(sp, machine_state_float_regset);
-        enter();
+  stmdb(sp, machine_state_regset);
+  vstmdb_f64(sp, machine_state_float_regset);
+  enter();
 }
 
 void MacroAssembler::restore_machine_state() {
-        leave();
-        vldmia_f64(sp, machine_state_float_regset);
-        ldmia(sp, machine_state_regset);
+  leave();
+  vldmia_f64(sp, machine_state_float_regset);
+  ldmia(sp, machine_state_regset);
 }
 
 void internal_internal_printf(const char *fmt, ...) {
@@ -2524,108 +2524,108 @@
 }
 
 void internal_printf(const char *format, uint32_t a, uint32_t b, uint32_t c) {
-        char buf[2048];
-        char fmt[2048];
-        buf[0] = '\0';
-        const char *thread_str = "THREAD 0x%08x : ";
-        int id = pthread_self();
-        strcpy(fmt, format);
-
-        char *str = strtok(fmt, "\n");
-        int nreplace = 0;
-        while(str) {
-                strcpy(buf, thread_str);
-                strcat(buf, str);
-                strcat(buf, "\n");
-                internal_internal_printf((const char*)buf, id, a, b, c);
-                str = strtok(NULL, "\n");
-        }
-}
-
-void MacroAssembler::get_bytecode(Register dst, Register bc) {
-        if(ENABLE_DEBUGGING) {
-                int nbytecodes = N_J_BYTECODES;
-                mov(dst, (address)j_bytecodes);
-                cmp(bc, nbytecodes);
-
-                ldr(dst, Address(dst, bc, lsl(2)), Assembler::LT);
-                ldr(dst, Address(dst, wordSize * nbytecodes), Assembler::GE);
+  char buf[2048];
+  char fmt[2048];
+  buf[0] = '\0';
+  const char *thread_str = "THREAD 0x%08x : ";
+  int id = pthread_self();
+  strcpy(fmt, format);
+
+  char *str = strtok(fmt, "\n");
+  int nreplace = 0;
+  while(str) {
+    strcpy(buf, thread_str);
+    strcat(buf, str);
+    strcat(buf, "\n");
+    internal_internal_printf((const char*)buf, id, a, b, c);
+    str = strtok(NULL, "\n");
   }
 }
 
+void MacroAssembler::get_bytecode(Register dst, Register bc) {
+  if(ENABLE_DEBUGGING) {
+    int nbytecodes = N_J_BYTECODES;
+    mov(dst, (address)j_bytecodes);
+    cmp(bc, nbytecodes);
+
+    ldr(dst, Address(dst, bc, lsl(2)), Assembler::LT);
+    ldr(dst, Address(dst, wordSize * nbytecodes), Assembler::GE);
+  }
+}
+
 int invocation_depth_count = -1; //TODO remove this with debugging info
 
 #define MAX_FCALL_DEPTH 4096
 struct thread_method_record{
-        int thread_id;
-        char names[MAX_FCALL_DEPTH][512];
-        int invocation_depth_count;
+  int thread_id;
+  char names[MAX_FCALL_DEPTH][512];
+  int invocation_depth_count;
 };
 int ntmrs = 0;
 #define MAX_TMRS 10
 thread_method_record tmr_list[MAX_TMRS];
 
 void push_tmr(Method *meth, int *thread_id, int *invocation_depth_count, char **name) {
-        int id = pthread_self();
-        *thread_id = id;
-        for(int i = 0; i < ntmrs; i++) {
-                thread_method_record *tmr = &tmr_list[i];
-                if(id == tmr->thread_id) {
-                        // Add a new frame
-                        if(tmr->invocation_depth_count >= -1 &&
-                           tmr->invocation_depth_count < (MAX_FCALL_DEPTH - 1)) {
-                                *invocation_depth_count = ++(tmr->invocation_depth_count);
-                                *name = tmr->names[tmr->invocation_depth_count];
-                                meth->name_and_sig_as_C_string(tmr->names[tmr->invocation_depth_count], 512);
-                                return;
-                        } else {
-                                fprintf(stderr, "%s : Invalid fcall depth index, %d\n", __FUNCTION__, tmr->invocation_depth_count);
-                                exit(1);
-                        }
-                }
-        }
-        // Add a new thread
-        if(ntmrs >= MAX_TMRS) {
-                fprintf(stderr, "Too many tmrs\n");
-                exit(1);
-        }
-        //Create a new tmr
-        tmr_list[ntmrs].thread_id = id;
-        tmr_list[ntmrs].invocation_depth_count = 0;
-        meth->name_and_sig_as_C_string(tmr_list[ntmrs].names[0], 512);
+  int id = pthread_self();
+  *thread_id = id;
+  for(int i = 0; i < ntmrs; i++) {
+    thread_method_record *tmr = &tmr_list[i];
+    if(id == tmr->thread_id) {
+      // Add a new frame
+      if(tmr->invocation_depth_count >= -1 &&
+        tmr->invocation_depth_count < (MAX_FCALL_DEPTH - 1)) {
+        *invocation_depth_count = ++(tmr->invocation_depth_count);
+        *name = tmr->names[tmr->invocation_depth_count];
+        meth->name_and_sig_as_C_string(tmr->names[tmr->invocation_depth_count], 512);
+        return;
+      } else {
+        fprintf(stderr, "%s : Invalid fcall depth index, %d\n", __FUNCTION__, tmr->invocation_depth_count);
+        exit(1);
+      }
+    }
+  }
+  // Add a new thread
+  if(ntmrs >= MAX_TMRS) {
+    fprintf(stderr, "Too many tmrs\n");
+    exit(1);
+  }
+  //Create a new tmr
+  tmr_list[ntmrs].thread_id = id;
+  tmr_list[ntmrs].invocation_depth_count = 0;
+  meth->name_and_sig_as_C_string(tmr_list[ntmrs].names[0], 512);
+  *invocation_depth_count = 0;
+  *name = tmr_list[ntmrs].names[0];
+  ntmrs++;
+}
+
+void pop_tmr(int *thread_id, int *invocation_depth_count, char **name) {
+  int id = pthread_self();
+  *thread_id = id;
+  for(int i = 0; i < ntmrs; i++) {
+    thread_method_record *tmr = &tmr_list[i];
+    if(id == tmr->thread_id) {
+      if(tmr->invocation_depth_count >= 0 &&
+        tmr->invocation_depth_count < MAX_FCALL_DEPTH) {
+        // Pop frame
+        *name = tmr->names[tmr->invocation_depth_count];
+        *invocation_depth_count = (tmr->invocation_depth_count)--;
+        return;
+      } else if ( -1 == tmr->invocation_depth_count) {
+        *name = (char*)"JVM-EXCEPTION-EXIT:(NOT-REALLY-A-FRAME)";
         *invocation_depth_count = 0;
-        *name = tmr_list[ntmrs].names[0];
-        ntmrs++;
+        return;
+      } else {
+        fprintf(stderr, "%s : Invalid fcall depth index, %d\n", __FUNCTION__, tmr->invocation_depth_count);
+        exit(1);
+      }
+    }
+  }
+  fprintf(stderr, "Unable to find suitable tmr\n");
+  exit(1);
 }
 
-void pop_tmr(int *thread_id, int *invocation_depth_count, char **name) {
-        int id = pthread_self();
-        *thread_id = id;
-        for(int i = 0; i < ntmrs; i++) {
-                thread_method_record *tmr = &tmr_list[i];
-                if(id == tmr->thread_id) {
-                        if(tmr->invocation_depth_count >= 0 &&
-                           tmr->invocation_depth_count < MAX_FCALL_DEPTH) {
-                                // Pop frame
-                                *name = tmr->names[tmr->invocation_depth_count];
-                                *invocation_depth_count = (tmr->invocation_depth_count)--;
-                                return;
-                        } else if ( -1 == tmr->invocation_depth_count) {
-                                *name = (char*)"JVM-EXCEPTION-EXIT:(NOT-REALLY-A-FRAME)";
-                                *invocation_depth_count = 0;
-                                return;
-                        } else {
-                                fprintf(stderr, "%s : Invalid fcall depth index, %d\n", __FUNCTION__, tmr->invocation_depth_count);
-                                exit(1);
-                        }
-                }
-        }
-        fprintf(stderr, "Unable to find suitable tmr\n");
-        exit(1);
-}
-
 void prepare_entry_exit_prefix(char *buf, int id, int invocation_depth_count) {
-        sprintf(buf, "THREAD 0x%08x : ", id);
+  sprintf(buf, "THREAD 0x%08x : ", id);
   for(int i = 0; i < invocation_depth_count; i++) {
     strcat(buf, "  ");
   }
@@ -2633,132 +2633,132 @@
 
 
 void print_entry(Method *meth, int native) {
-        char *name;
-        int invocation_depth_count, id;
-        push_tmr(meth, &id, &invocation_depth_count, &name);
-
-        if(MacroAssembler::enable_method_debug) {
-                char buf[4096], buf_b[2048];
-                prepare_entry_exit_prefix(buf, id, invocation_depth_count);
-                if(native) {
-                  sprintf(buf_b, "CALL NATIVE : %s\n", name);
-                } else {
-                  sprintf(buf_b, "CALL JAVA   : %s\n", name);
-                }
-                strcat(buf, buf_b);
-                printf("%s", buf);
-                fflush(stdout);
-        }
+  char *name;
+  int invocation_depth_count, id;
+  push_tmr(meth, &id, &invocation_depth_count, &name);
+
+  if(MacroAssembler::enable_method_debug) {
+    char buf[4096], buf_b[2048];
+    prepare_entry_exit_prefix(buf, id, invocation_depth_count);
+    if(native) {
+      sprintf(buf_b, "CALL NATIVE : %s\n", name);
+    } else {
+      sprintf(buf_b, "CALL JAVA   : %s\n", name);
+    }
+    strcat(buf, buf_b);
+    printf("%s", buf);
+    fflush(stdout);
+  }
 }
 
 void print_exit(bool normal) {
-        char *name;
-        int invocation_depth_count, id;
-        pop_tmr(&id, &invocation_depth_count, &name);
-
-        if(MacroAssembler::enable_method_debug) {
-                char buf[4096], buf_b[2048];
-                prepare_entry_exit_prefix(buf, id, invocation_depth_count);
-                sprintf(buf_b, normal ? "EXIT        : %s\n" : "EXCPN EXIT  : %s\n", name);
-                strcat(buf, buf_b);
-                printf("%s", buf);
-                fflush(stdout);
-        }
+  char *name;
+  int invocation_depth_count, id;
+  pop_tmr(&id, &invocation_depth_count, &name);
+
+  if(MacroAssembler::enable_method_debug) {
+    char buf[4096], buf_b[2048];
+    prepare_entry_exit_prefix(buf, id, invocation_depth_count);
+    sprintf(buf_b, normal ? "EXIT        : %s\n" : "EXCPN EXIT  : %s\n", name);
+    strcat(buf, buf_b);
+    printf("%s", buf);
+    fflush(stdout);
+  }
 }
 
 void MacroAssembler::print_method_entry(Register rmethod, bool native) {
-        if(ENABLE_DEBUGGING) {
-                save_machine_state();
-
-                bic(sp, sp, 7); // 8-byte align stack
-                mov(rscratch2, (address)print_entry);
-                mov(r0, rmethod);
-                mov(r1, native);
-                bl(rscratch2);
-
-                restore_machine_state();
-        }
+  if(ENABLE_DEBUGGING) {
+    save_machine_state();
+
+    bic(sp, sp, 7); // 8-byte align stack
+    mov(rscratch2, (address)print_entry);
+    mov(r0, rmethod);
+    mov(r1, native);
+    bl(rscratch2);
+
+    restore_machine_state();
+  }
 }
 
 void MacroAssembler::print_method_exit(bool normal) {
-        if(ENABLE_DEBUGGING) {
-                save_machine_state();
-
-                bic(sp, sp, 7); // 8-byte align stack
-                mov(rscratch2, (address)print_exit);
-                mov(r0, normal);
-                bl(rscratch2);
-
-                restore_machine_state();
-        }
+  if(ENABLE_DEBUGGING) {
+    save_machine_state();
+
+    bic(sp, sp, 7); // 8-byte align stack
+    mov(rscratch2, (address)print_exit);
+    mov(r0, normal);
+    bl(rscratch2);
+
+    restore_machine_state();
+  }
 }
 
 void MacroAssembler::reg_printf_internal(bool important, const char *fmt, Register ra, Register rb, Register rc) {
-        if(ENABLE_DEBUGGING) {
-                Label skip;
-                save_machine_state();
-
-                str(ra, Address(pre(sp, -wordSize)));
-                str(rb, Address(pre(sp, -wordSize)));
-                str(rc, Address(pre(sp, -wordSize)));
-
-                if(!important) {
-                        mov(r0, (address)&enable_debug);
-                        ldr(r0, Address(r0));
-                        cmp(r0, 0);
-                        b(skip, Assembler::EQ);
-                }
-
-                int sp_difference = wordSize * (count_bits(machine_state_regset) +
-                                                2 * count_bits(machine_state_float_regset) +
-                                                2 + 3); //Frame entry and saved
-
-                mov(r0, (address)fmt);
-                if(ra != sp) ldr(r1, Address(sp, 2 * wordSize));
-                else         add(r1, sp, sp_difference);
-
-                if(rb != sp) ldr(r2, Address(sp, wordSize));
-                else         add(r2, sp, sp_difference);
-
-                if(rc != sp) ldr(r3, Address(sp));
-                else         add(r3, sp, sp_difference);
-
-                bic(sp, sp, 7); // 8-byte align stack
-
-                mov(rscratch2, (address)internal_printf);
-                bl(rscratch2);
-
-                bind(skip);
-                restore_machine_state();
-        }
+  if(ENABLE_DEBUGGING) {
+    Label skip;
+    save_machine_state();
+
+    str(ra, Address(pre(sp, -wordSize)));
+    str(rb, Address(pre(sp, -wordSize)));
+    str(rc, Address(pre(sp, -wordSize)));
+
+    if(!important) {
+      mov(r0, (address)&enable_debug);
+      ldr(r0, Address(r0));
+      cmp(r0, 0);
+      b(skip, Assembler::EQ);
+    }
+
+    int sp_difference = wordSize * (count_bits(machine_state_regset) +
+                                    2 * count_bits(machine_state_float_regset) +
+                                    2 + 3); //Frame entry and saved
+
+    mov(r0, (address)fmt);
+    if(ra != sp) ldr(r1, Address(sp, 2 * wordSize));
+    else         add(r1, sp, sp_difference);
+
+    if(rb != sp) ldr(r2, Address(sp, wordSize));
+    else         add(r2, sp, sp_difference);
+
+    if(rc != sp) ldr(r3, Address(sp));
+    else         add(r3, sp, sp_difference);
+
+    bic(sp, sp, 7); // 8-byte align stack
+
+    mov(rscratch2, (address)internal_printf);
+    bl(rscratch2);
+
+    bind(skip);
+    restore_machine_state();
+  }
 }
 
 void MacroAssembler::reg_printf(const char *fmt, Register ra, Register rb, Register rc) {
-        reg_printf_internal(false, fmt, ra, rb, rc);
+  reg_printf_internal(false, fmt, ra, rb, rc);
 }
 
 void MacroAssembler::reg_printf_important(const char *fmt, Register ra, Register rb, Register rc) {
-        reg_printf_internal(true, fmt, ra, rb, rc);
+  reg_printf_internal(true, fmt, ra, rb, rc);
 }
 
 // When debugging, set the break on bkpnt
 void bkpnt() { return; }
 void MacroAssembler::create_breakpoint() {
-        if(ENABLE_DEBUGGING) {
-                save_machine_state();
-                bic(sp, sp, 7); // 8-byte align stack
-
-                mov(rscratch2, (address) bkpnt);
-                bl(rscratch2);
-
-                restore_machine_state();
-        }
+  if(ENABLE_DEBUGGING) {
+    save_machine_state();
+    bic(sp, sp, 7); // 8-byte align stack
+
+    mov(rscratch2, (address) bkpnt);
+    bl(rscratch2);
+
+    restore_machine_state();
+  }
 }
 
 
 void MacroAssembler::print_cpool(InstanceKlass *klass) {
-        ttyLocker ttyl;
-        klass->constants()->print_on(tty);
+  ttyLocker ttyl;
+  klass->constants()->print_on(tty);
 }
 
 
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/macroAssembler_aarch32.hpp
--- a/src/cpu/aarch32/vm/macroAssembler_aarch32.hpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/macroAssembler_aarch32.hpp	Fri Jan 15 23:23:56 2016 +0800
@@ -92,7 +92,7 @@
   void call_VM_helper(Register oop_result, address entry_point, int number_of_arguments, bool check_exceptions = true);
 
  public:
-        void init_unseen_bytecodes();
+  void init_unseen_bytecodes();
   MacroAssembler(CodeBuffer* code) : Assembler(code) { init_unseen_bytecodes();}
 
   // Biased locking support
@@ -690,33 +690,33 @@
   //address read_polling_page(Register r, relocInfo::relocType rtype);
 
 
-        // Auto dispatch for barriers isb, dmb & dsb.
-        void isb() {
-                if(VM_Version::features() & FT_ARMV7) {
-                        Assembler::isb();
-                } else {
-                        cp15isb();
-                }
-        }
+  // Auto dispatch for barriers isb, dmb & dsb.
+  void isb() {
+    if(VM_Version::features() & FT_ARMV7) {
+      Assembler::isb();
+    } else {
+      cp15isb();
+    }
+  }
 
-        void dsb(enum barrier option) {
-                if(VM_Version::features() & FT_ARMV7) {
-                        Assembler::dsb(option);
-                } else {
-                        cp15dsb();
-                }
+  void dsb(enum barrier option) {
+    if(VM_Version::features() & FT_ARMV7) {
+      Assembler::dsb(option);
+    } else {
+      cp15dsb();
+    }
   }
 
   void dmb(enum barrier option) {
-                if(VM_Version::features() & FT_ARMV7) {
-                        Assembler::dmb(option);
-                } else {
-                        cp15dmb();
-                }
+    if(VM_Version::features() & FT_ARMV7) {
+      Assembler::dmb(option);
+    } else {
+      cp15dmb();
+    }
   }
 
   void membar(Membar_mask_bits order_constraint) {
-          dmb(Assembler::barrier(order_constraint));
+    dmb(Assembler::barrier(order_constraint));
   }
 
   // ISB may be needed because of a safepoint
@@ -727,13 +727,13 @@
   void mult_long(Register Rd, Register Rn, Register Rm);
 
  private:
-        void divide32(Register res, Register num, Register den, bool want_mod);
+  void divide32(Register res, Register num, Register den, bool want_mod);
  public:
-        // <Rd+1:Rd> = <Rn+1:Rn> / <Rm+1:Rm>
-        // <Rd+1:Rd> = <Rn+1:Rn> % <Rm+1:Rm>
-        // <Rd> = <Rn> / <Rm>
-        // <Rd> = <Rn> % <Rm>
-        void divide(Register Rd, Register Rn, Register Rm, int width, bool want_remainder);
+  // <Rd+1:Rd> = <Rn+1:Rn> / <Rm+1:Rm>
+  // <Rd+1:Rd> = <Rn+1:Rn> % <Rm+1:Rm>
+  // <Rd> = <Rn> / <Rm>
+  // <Rd> = <Rn> % <Rm>
+  void divide(Register Rd, Register Rn, Register Rm, int width, bool want_remainder);
 
   void extract_bits(Register dest, Register source, int lsb, int width);
 
@@ -750,8 +750,8 @@
   void save_machine_state();
   void restore_machine_state();
 
-        static uint32_t bytecodes_until_print;
-        static uint32_t bytecodes_executed;
+  static uint32_t bytecodes_until_print;
+  static uint32_t bytecodes_executed;
   static int enable_debug;
   static int enable_method_debug;
   static int enable_debugging_static;
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/sharedRuntime_aarch32.cpp
--- a/src/cpu/aarch32/vm/sharedRuntime_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/sharedRuntime_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -231,7 +231,7 @@
                                            int total_args_passed,
                                            int is_outgoing) {
 
-        // FIXME This looks like it needs to be fixed
+  // FIXME This looks like it needs to be fixed
   // Create the mapping between argument positions and
   // registers.
   static const Register INT_ArgReg[Argument::n_int_register_parameters_j] = {
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/stubGenerator_aarch32.cpp
--- a/src/cpu/aarch32/vm/stubGenerator_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/stubGenerator_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -241,17 +241,17 @@
 #endif
 
     if(MacroAssembler::enable_debugging_static){
-        // FIXME Remove this hacky debugging code
-        Label L;
+      // FIXME Remove this hacky debugging code
+      Label L;
       __ ldr(rscratch2, Address(rthread, Thread::pending_exception_offset()));
-        __ cbnz(rscratch2, L);
-        // If we're returning via an exception then we shouldn't report exit,
-        // the exception handler will have already reported the exit and reporting
-        // via our progress through the call stub will result in an extra method
-        // being reported as exited.
-            __ print_method_exit();
-            __ bind(L);
-          }
+      __ cbnz(rscratch2, L);
+      // If we're returning via an exception then we shouldn't report exit,
+      // the exception handler will have already reported the exit and reporting
+      // via our progress through the call stub will result in an extra method
+      // being reported as exited.
+      __ print_method_exit();
+      __ bind(L);
+    }
 
     // NOTE Horrible tricks here
     // We saved c_rarg0, c_rarg1, c_rarg2, however we don't need to restore
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/templateInterpreter_aarch32.cpp
--- a/src/cpu/aarch32/vm/templateInterpreter_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/templateInterpreter_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -653,7 +653,7 @@
   // point rfp to location of old pc
   __ add(rfp, sp, 9 * wordSize);
 
-        __ reg_printf("Three-quarters through\n");
+  __ reg_printf("Three-quarters through\n");
   // set sender sp
   // leave last_sp as null
   __ mov(rscratch1, 0);
@@ -946,11 +946,11 @@
   // However, large signatures cannot be cached and are generated
   // each time here.  The slow-path generator can do a GC on return,
   // so we must reload it after the call.
-        __ reg_printf("**BEFORE**\nrlocals = %p,locals_esp = %p, sp = %p\n", rlocals, locals_esp, sp);
-        __ reg_printf("About to call the Method::signature_handler = %p\n", rscratch1);
+  __ reg_printf("**BEFORE**\nrlocals = %p,locals_esp = %p, sp = %p\n", rlocals, locals_esp, sp);
+  __ reg_printf("About to call the Method::signature_handler = %p\n", rscratch1);
   __ bl(rscratch1);
   __ reg_printf("**AFER**\nr0 : %p, r1 : %p, r2 : %p\n", r0, r1, r2);
-        __ reg_printf("r3 : %p, sp : %p\n", r3, sp);
+  __ reg_printf("r3 : %p, sp : %p\n", r3, sp);
   __ get_method(rmethod);        // slow path can do a GC, reload rmethod
 
 
@@ -1025,8 +1025,8 @@
   __ reg_printf("Calling native method, lr = %p & rmethod = %p\n", lr, rmethod);
   // Call the native method.
   /*__ reg_printf("**ONCALL**\nr0 : %p\nr1 : %p\nr2 : %p\n", r0, r1, r2);
-        __ reg_printf("r3 : %p\n\nr4 : %p\nrloc : %p\n", r3, r4, rlocals);*/
-        __ reg_printf("Stack Pointer on entry to native, sp = %p\n", sp);
+  __ reg_printf("r3 : %p\n\nr4 : %p\nrloc : %p\n", r3, r4, rlocals);*/
+  __ reg_printf("Stack Pointer on entry to native, sp = %p\n", sp);
   __ bl(native_entry_point);
   __ reg_printf("Returned from native, lr = %p, r1 = %p, r0 = %p\n", lr, r1, r0);
   __ maybe_isb();
@@ -1813,7 +1813,7 @@
 
   Interpreter::_remove_activation_entry = __ pc();
   __ print_method_exit(false);
-        __ reg_printf("remove_activation_entry\n");
+  __ reg_printf("remove_activation_entry\n");
 
   // preserve exception over this code sequence
   __ pop_ptr(r0);
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/templateTable_aarch32.cpp
--- a/src/cpu/aarch32/vm/templateTable_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/templateTable_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -82,7 +82,7 @@
 }
 
 static inline Address iaddress(Register r) {
-        // FIXME
+  // FIXME
   return Address(rlocals, r, lsl(2));
 }
 
@@ -341,11 +341,11 @@
   float fval = value;
   assert(value == 0 || value == 1 || value == 2, "invalid float const");
   if(VM_Version::features() & FT_VFPV3) {
-                __ vmov_f32(d0, fval);
-        } else {
-                __ mov(r0, *((uint32_t*)&fval));
-                __ vmov_f32(d0, r0);
-        }
+    __ vmov_f32(d0, fval);
+  } else {
+    __ mov(r0, *((uint32_t*)&fval));
+    __ vmov_f32(d0, r0);
+  }
 }
 
 void TemplateTable::dconst(int value)
@@ -354,13 +354,13 @@
   double dval = value;
   assert(value == 0 || value == 1 || value == 2, "invalid double const");
   if(VM_Version::features() & FT_VFPV3) {
-                __ vmov_f64(d0, dval);
-        } else {
-                uint32_t* ptr = (uint32_t*)&dval;
-                __ mov(r0, *ptr);
-                __ mov(r1, *(ptr + 1));
-                __ vmov_f64(d0, r0, r1);
-        }
+    __ vmov_f64(d0, dval);
+  } else {
+    uint32_t* ptr = (uint32_t*)&dval;
+    __ mov(r0, *ptr);
+    __ mov(r1, *(ptr + 1));
+    __ vmov_f64(d0, r0, r1);
+  }
 }
 
 void TemplateTable::bipush()
@@ -1354,14 +1354,14 @@
   // shift count is in r0 - take shift from bottom six bits only
   __ andr(r0, r0, 63);
   __ pop_l(r2); //LSB in lowest reg
-        int word_bytes = 8 * wordSize;
-
-        __ sub(r1, r0, word_bytes);
-        __ lsl(r3, r3, r0);
-        __ orr(r3, r3, r2, lsl(r1));
-        __ rsb(r1, r0, word_bytes);
-        __ orr(r1, r3, r2, lsr(r1));
-        __ lsl(r0, r2, r0);
+  int word_bytes = 8 * wordSize;
+
+  __ sub(r1, r0, word_bytes);
+  __ lsl(r3, r3, r0);
+  __ orr(r3, r3, r2, lsl(r1));
+  __ rsb(r1, r0, word_bytes);
+  __ orr(r1, r3, r2, lsr(r1));
+  __ lsl(r0, r2, r0);
 }
 
 void TemplateTable::lshr()
@@ -1420,10 +1420,10 @@
     __ vcvt_f64_f32(d1, d0);
     __ pop_f(d0);
     __ vcvt_f64_f32(d0, d0);
-    #ifndef HARD_FLOAT_CC
+#ifndef HARD_FLOAT_CC
     __ vmov_f64(r0, r1, d0);
     __ vmov_f64(r2, r3, d1);
-    #endif
+#endif
     __ mov(rscratch1, (address)fmod);
     __ bl(rscratch1);
     __ vcvt_f32_f64(d0, d0);
@@ -1457,10 +1457,10 @@
   case rem:
     __ vmov_f64(d1, d0);
     __ pop_d(d0);
-    #ifndef HARD_FLOAT_CC
+#ifndef HARD_FLOAT_CC
     __ vmov_f64(r0, r1, d0);
     __ vmov_f64(r2, r3, d1);
-    #endif
+#endif
     __ mov(rscratch1, (address)fmod);
     __ bl(rscratch1);
     break;
@@ -1579,7 +1579,7 @@
     __ reg_printf("Convert i2l (after) 0x%08x%08x\n", r1, r0);
     break;
   case Bytecodes::_i2f:
-        //__ bkpt(735);
+    //__ bkpt(735);
     //__ scvtfws(d0, r0);
     //__ reg_printf("VCVT Convert i2f, (before) 0x%08x\n", r0);
     __ vmov_f32(d0, r0);
@@ -1635,10 +1635,10 @@
   case Bytecodes::_f2l:
   {
     //float already in d0 long goes to <r1:r0>
-    #ifndef HARD_FLOAT_CC
+#ifndef HARD_FLOAT_CC
     //Need to move float in d0 to r0
     __ vmov_f32(r0, d0);
-    #endif
+#endif
     __ call_VM_leaf_base(CAST_FROM_FN_PTR(address, SharedRuntime::f2l), 0);
   }
     break;
@@ -1665,10 +1665,10 @@
   case Bytecodes::_d2l:
   {
     // d0 -> <r1:r0>
-    #ifndef HARD_FLOAT_CC
+#ifndef HARD_FLOAT_CC
     //Need to move float in d0 to r0
     __ vmov_f64(r0, r1, d0);
-    #endif
+#endif
     __ call_VM_leaf_base(CAST_FROM_FN_PTR(address, SharedRuntime::d2l), 0);
   }
     break;
@@ -1713,7 +1713,7 @@
 
 void TemplateTable::float_cmp(bool is_float, int unordered_result)
 {
-        //__ bkpt(400);
+  //__ bkpt(400);
   if (is_float) {
     __ pop_f(d1);
     __ vcmp_f32(d1, d0);
@@ -2113,7 +2113,7 @@
   transition(itos, vtos);
   Label loop_entry, loop, found, continue_execution;
 
-        __ reg_printf("Linearswitching to value %d\n", r0);
+  __ reg_printf("Linearswitching to value %d\n", r0);
 
   // bswap r0 so we can avoid bswapping the table entries
   __ rev(r0, r0);
@@ -2526,7 +2526,7 @@
   Label Done, notByte, notInt, notShort, notChar,
               notLong, notFloat, notObj, notDouble;
 
-        //__ bkpt(324);
+  //__ bkpt(324);
   // x86 uses a shift and mask or wings it with a shift plus assert
   // the mask is not needed. aarch32 just uses bitfield extract
   __ extract_bits(flags, flags, ConstantPoolCacheEntry::tos_state_shift,  ConstantPoolCacheEntry::tos_state_bits);
@@ -2596,7 +2596,7 @@
   __ cmp(flags, ltos);
   __ b(notLong, Assembler::NE);
   // ltos
-        __ lea(rscratch1, field);
+  __ lea(rscratch1, field);
   __ atomic_ldrd(r0, r1, rscratch1);
   __ push(ltos);
   // Rewrite bytecode to be faster
@@ -3006,7 +3006,7 @@
     do_oop_store(_masm, field, r0, _bs->kind(), false);
     break;
   case Bytecodes::_fast_lputfield:
-                __ lea(rscratch1, field);
+    __ lea(rscratch1, field);
     __ atomic_strd(r0, r1, rscratch1, r2, r3);
     break;
   case Bytecodes::_fast_iputfield:
@@ -3025,8 +3025,8 @@
     __ vstr_f32(d0, Address(rscratch1));
     break;
   case Bytecodes::_fast_dputfield:
-                __ lea(rscratch1, field);
-                __ vmov_f64(r0, r1, d0);
+    __ lea(rscratch1, field);
+    __ vmov_f64(r0, r1, d0);
     __ atomic_strd(r0, r1, rscratch1, r2, r3);
     break;
   default:
@@ -3107,10 +3107,10 @@
     __ lea(r0, field); // r0 <= field
     __ vldr_f32(d0, Address(r0));
     __ vmov_f32(rscratch1, d0);
-           break;
+    break;
   case Bytecodes::_fast_dgetfield:
     __ lea(rscratch1, field); // r0 <= field
-                __ atomic_ldrd(r0, r1, rscratch1);
+    __ atomic_ldrd(r0, r1, rscratch1);
     __ vmov_f64(d0, r0, r1);
     break;
   default:
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/vm_version_aarch32.cpp
--- a/src/cpu/aarch32/vm/vm_version_aarch32.cpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/vm_version_aarch32.cpp	Fri Jan 15 23:23:56 2016 +0800
@@ -97,19 +97,19 @@
 
 
 bool VM_Version::identify_procline(const char *tag, char **line) {
-        char *i = *line;
-        const char EOT = '\t', EOT2 = ':'; // the longest has no tabs
-        for(; '\0' != *i && EOT != *i && EOT2 != *i; i++);
-        if(EOT == *i || EOT2 == *i) {
-                if(!memcmp(*line, tag, i - *line)) {
-                        for(i++; (EOT == *i || EOT2 == *i || ' ' == *i) && '\0' != *i; i++);
-                        if('\0' != *i) {
-                                *line = i;
-                                return true;
-                        }
-                }
-        }
-        return false;
+  char *i = *line;
+  const char EOT = '\t', EOT2 = ':'; // the longest has no tabs
+  for(; '\0' != *i && EOT != *i && EOT2 != *i; i++);
+  if(EOT == *i || EOT2 == *i) {
+    if(!memcmp(*line, tag, i - *line)) {
+      for(i++; (EOT == *i || EOT2 == *i || ' ' == *i) && '\0' != *i; i++);
+      if('\0' != *i) {
+        *line = i;
+        return true;
+      }
+    }
+  }
+  return false;
 }
 
 void VM_Version::get_processor_features() {
@@ -129,67 +129,67 @@
   FLAG_SET_DEFAULT(UseSSE42Intrinsics, false);
 
   int ncores = 0, cpu, variant, model, revision;
-        enum ProcessorFeatures f = FT_NONE;
-        char buf[2048], *i;
-        if(FILE *fp = fopen("/proc/cpuinfo", "r")) {
-                while((i = fgets(buf, 2048, fp))) {
-                        if(identify_procline("Features", &i)) {
-                                i = strtok(i, " \n");
-                                while(i) {
-                                        if(!strcmp("idiva", i)) {
-                                                f = (ProcessorFeatures)(f | FT_HW_DIVIDE);
-                                        } else if(!strcmp("vfpv3", i) || !strcmp("vfpv4", i)) {
-                                                // Assuming that vfpv4 implements all of vfpv3
-                                                // and that they both implement all of v2.
-                                                f = (ProcessorFeatures)(f | FT_VFPV3 | FT_VFPV2);
-                                        } else if(!strcmp("vfp", i)) {
-                                                // Assuming that VFPv2 is identified by plain vfp
-                                                f = (ProcessorFeatures)(f | FT_VFPV2);
-                                        }
-                                        i = strtok(NULL, " \n");
-                                }
-                        } else if(identify_procline("Processor", &i)) {
-                                i = strtok(i, " \n");
-                                while(i) {
-                                        // if the info is read correctly do
-                                        if(!strcmp("ARMv7", i) || !strcmp("AArch64", i)) {
-                                                f = (ProcessorFeatures)(f | FT_ARMV7);
-                                        } else if(!strcmp("ARMv6-compatible", i)) {
-                                                //TODO sort out the ARMv6 identification code
-                                        }
-                                        i = strtok(NULL, " \n");
-                                }
-                        } else if(identify_procline("model name", &i)) {
-                                i = strtok(i, " \n");
-                                while(i) {
-                                        // if the info is read correctly do
-                                        if(!strcmp("ARMv7", i)) {
-                                                f = (ProcessorFeatures)(f | FT_ARMV7);
-                                        } else if(!strcmp("ARMv6-compatible", i)) {
-                                                //TODO sort out the ARMv6 identification code
-                                        }
-                                        i = strtok(NULL, " \n");
-                                }
-                        } else if(identify_procline("processor", &i)) {
-                                ncores++;
-                        } else if(identify_procline("CPU implementer", &i)) {
-                                cpu = strtol(i, NULL, 0);
-                        } else if(identify_procline("CPU variant", &i)) {
-                                variant = strtol(i, NULL, 0);
-                        } else if(identify_procline("CPU part", &i)) {
-                                model = strtol(i, NULL, 0);
-                        } else if(identify_procline("CPU revision", &i)) {
-                                revision = strtol(i, NULL, 0);
-                        }
-                }
-                fclose(fp);
+  enum ProcessorFeatures f = FT_NONE;
+  char buf[2048], *i;
+  if(FILE *fp = fopen("/proc/cpuinfo", "r")) {
+    while((i = fgets(buf, 2048, fp))) {
+      if(identify_procline("Features", &i)) {
+        i = strtok(i, " \n");
+        while(i) {
+          if(!strcmp("idiva", i)) {
+            f = (ProcessorFeatures)(f | FT_HW_DIVIDE);
+          } else if(!strcmp("vfpv3", i) || !strcmp("vfpv4", i)) {
+            // Assuming that vfpv4 implements all of vfpv3
+            // and that they both implement all of v2.
+            f = (ProcessorFeatures)(f | FT_VFPV3 | FT_VFPV2);
+          } else if(!strcmp("vfp", i)) {
+            // Assuming that VFPv2 is identified by plain vfp
+            f = (ProcessorFeatures)(f | FT_VFPV2);
+          }
+          i = strtok(NULL, " \n");
         }
-        if(1 == ncores) {
-                f = (ProcessorFeatures)(f | FT_SINGLE_CORE);
+      } else if(identify_procline("Processor", &i)) {
+        i = strtok(i, " \n");
+        while(i) {
+          // if the info is read correctly do
+          if(!strcmp("ARMv7", i) || !strcmp("AArch64", i)) {
+            f = (ProcessorFeatures)(f | FT_ARMV7);
+          } else if(!strcmp("ARMv6-compatible", i)) {
+            //TODO sort out the ARMv6 identification code
+          }
+          i = strtok(NULL, " \n");
         }
-        _features = f;
-        sprintf(buf, "0x%02x:0x%x:0x%03x:%d", cpu, variant, model, revision);
-        _cpu_features = os::strdup(buf);
+      } else if(identify_procline("model name", &i)) {
+        i = strtok(i, " \n");
+        while(i) {
+          // if the info is read correctly do
+          if(!strcmp("ARMv7", i)) {
+            f = (ProcessorFeatures)(f | FT_ARMV7);
+          } else if(!strcmp("ARMv6-compatible", i)) {
+            //TODO sort out the ARMv6 identification code
+          }
+          i = strtok(NULL, " \n");
+        }
+      } else if(identify_procline("processor", &i)) {
+        ncores++;
+      } else if(identify_procline("CPU implementer", &i)) {
+        cpu = strtol(i, NULL, 0);
+      } else if(identify_procline("CPU variant", &i)) {
+        variant = strtol(i, NULL, 0);
+      } else if(identify_procline("CPU part", &i)) {
+        model = strtol(i, NULL, 0);
+      } else if(identify_procline("CPU revision", &i)) {
+        revision = strtol(i, NULL, 0);
+      }
+    }
+    fclose(fp);
+  }
+  if(1 == ncores) {
+    f = (ProcessorFeatures)(f | FT_SINGLE_CORE);
+  }
+  _features = f;
+  sprintf(buf, "0x%02x:0x%x:0x%03x:%d", cpu, variant, model, revision);
+  _cpu_features = os::strdup(buf);
 
 #ifdef COMPILER2
   if (UseMultiplyToLenIntrinsic) {
@@ -205,10 +205,10 @@
   }*/
 
   /*if(!(f & FT_ARMV7) && FLAG_IS_DEFAULT(UseMembar)) {
-        UseMembar = false;
+  UseMembar = false;
   } else if(UseMembar) {
-        fprintf(stderr, "Unable to use memory barriers as not on ARMv7, disabling.\n");
-        UseMembar = false;
+  fprintf(stderr, "Unable to use memory barriers as not on ARMv7, disabling.\n");
+  UseMembar = false;
   }*/
 }
 
diff -r dc9b82a9fbc7 src/cpu/aarch32/vm/vm_version_aarch32.hpp
--- a/src/cpu/aarch32/vm/vm_version_aarch32.hpp	Mon Jan 11 14:58:34 2016 +0000
+++ b/src/cpu/aarch32/vm/vm_version_aarch32.hpp	Fri Jan 15 23:23:56 2016 +0800
@@ -31,19 +31,19 @@
 #include "runtime/vm_version.hpp"
 
 enum ProcessorFeatures {
-        FT_NONE = 0,
-        FT_HW_DIVIDE = 1,
-        FT_VFPV2 = 2,
-        FT_VFPV3 = 4,
-        FT_ARMV7 = 8,
-        FT_ARMV6T2 = 16,
-        FT_ARMV6K = 32,
-        FT_SINGLE_CORE = 64
+  FT_NONE = 0,
+  FT_HW_DIVIDE = 1,
+  FT_VFPV2 = 2,
+  FT_VFPV3 = 4,
+  FT_ARMV7 = 8,
+  FT_ARMV6T2 = 16,
+  FT_ARMV6K = 32,
+  FT_SINGLE_CORE = 64
 };
 
 class VM_Version : public Abstract_VM_Version {
  public:
-        // Processor feature lookup.
+  // Processor feature lookup.
 
   enum {
     CPU_ARM       = 'A',
@@ -63,17 +63,17 @@
   static void initialize();
 
  private:
-        static enum ProcessorFeatures _features;
-        static const char* _cpu_features;
+  static enum ProcessorFeatures _features;
+  static const char* _cpu_features;
 
   static void get_processor_features();
-        static bool identify_procline(const char *tag, char **line);
+  static bool identify_procline(const char *tag, char **line);
 
  public:
-        static enum ProcessorFeatures features() {
-                return _features;
-        }
-        static const char* cpu_features() { return _cpu_features; }
+  static enum ProcessorFeatures features() {
+    return _features;
+  }
+  static const char* cpu_features() { return _cpu_features; }
 };
 
 #endif // CPU_AARCH32_VM_VM_VERSION_AARCH32_HPP


More information about the aarch32-port-dev mailing list