ARMv5 with Valgrind: illegal instruction - arm

thanks to other posts here, I managed to build and install Valgrind for ARM:
make clean
make distclean
export PATH=$PATH:/usr/local/angstrom/arm/bin
export CROSS_COMPILE=/usr/local/angstrom/arm/bin/arm-angstrom-linux-gnueabi-
export CC=${CROSS_COMPILE}gcc
export CPP=${CROSS_COMPILE}cpp
export CXX=${CROSS_COMPILE}g++
export LD=${CROSS_COMPILE}ld
export AR=${CROSS_COMPILE}ar
export RANLIB=${CROSS_COMPILE}ranlib
sed -i -e "s#armv7#arm#g" configure
./configure --host=arm-none-linux-gnueabi \
--host=arm-none-linux-gnueabi \
--prefix=/opt/valgrind/ \
CFLAGS=-static
make
make install
and copied it into the same path on the ARMv5 machine.
I am running an ARMv5 and apparently Valgrind is still not supporting ARMv5.
When starting Valgrind, I'll get
Illegal instruction
and I suppose this is, because Valgrind needs the ARMv7 instruction set?
So:
did I make any error in configure when cross-compiling valgrind?
is there any way to get Valgrind up and running on ARMv5 or even only the vgdb or Memcheck tool included in Valgrind?
Is there any other tool to do memchecks on ARMv5, I have heard about
Adress Sanitizer or
Malloc_Check
and are they recommendable for ARMv5?
Thanks a lot for help and support!
BR Florian
I have created a core dump and run gdb, unfortunately without debugging symbols:
What command should i used?
[New Thread 10495]
Core was generated by `/opt/valgrind/bin/valgrind'.
Program terminated with signal 4, Illegal instruction.
#0 0x38035804 in ?? ()
(gdb) list
305 platform ? platform : "unknown");
306
307 return platform;
308 }
309
310 /* Where we expect to find all our aux files */
311 static const char *valgrind_lib = VG_LIBDIR;
312
313 int main(int argc, char** argv, char** envp)
314 {
(gdb) info frame
Stack level 0, frame at 0x38ac7894:
pc = 0x38035804; saved pc 0x380425e0
called by frame at 0x38ac7894
Cannot access memory at address 0x0
Arglist at 0x38ac7894, args:
Locals at 0x38ac7894, Previous frame's sp is 0x38ac7894
(gdb) list main
326
327 /* Start the debugging-log system ASAP. First find out how many
328 "-d"s were specified. This is a pre-scan of the command line.
329 At the same time, look for the tool name. */
330 loglevel = 0;
331 for (i = 1; i < argc; i++) {
332 if (argv[i][0] != '-') {
333 clientname = argv[i];
334 break;
335 }
(gdb) disassemble main
Dump of assembler code for function main:
0x000087b4 <+0>: cmp r0, #1
0x000087b8 <+4>: push {r4, r5, r6, r7, r8, r9, r10, r11, lr}
0x000087bc <+8>: mov r10, r0
0x000087c0 <+12>: sub sp, sp, #12
0x000087c4 <+16>: mov r11, r1
0x000087c8 <+20>: mov r9, r2
0x000087cc <+24>: ble 0x8c90 <main+1244>
0x000087d0 <+28>: ldr r4, [r1, #4]
0x000087d4 <+32>: ldrb r3, [r4]
0x000087d8 <+36>: cmp r3, #45 ; 0x2d
0x000087dc <+40>: movne r8, #0
0x000087e0 <+44>: movne r7, r8
0x000087e4 <+48>: bne 0x8820 <main+108>
0x000087e8 <+52>: ldr r0, [pc, #1372] ; 0x8d4c <main+1432>
0x000087ec <+56>: bl 0x29d20 <strlen>
0x000087f0 <+60>: mov r8, #0
0x000087f4 <+64>: mov r5, #1
0x000087f8 <+68>: mov r7, r8
0x000087fc <+72>: mov r6, r0
0x00008800 <+76>: cmp r6, #3
0x00008804 <+80>: bhi 0x8b70 <main+956>
0x00008808 <+84>: cmp r6, #0
0x0000880c <+88>: bne 0x8a00 <main+588>
0x00008810 <+92>: add r0, r5, #1
0x00008814 <+96>: cmp r10, r0
0x00008818 <+100>: ble 0x8b90 <main+988>
0x0000881c <+104>: ldr r4, [r11, r0, lsl #2]
0x00008820 <+108>: mov r0, r8
0x00008824 <+112>: ldr r1, [pc, #1316] ; 0x8d50 <main+1436>
0x00008828 <+116>: bl 0xa258 <vgPlain_debugLog_startup>
0x0000882c <+120>: cmp r7, #0
0x00008830 <+124>: mov r10, r7
0x00008834 <+128>: beq 0x8bac <main+1016>
0x00008838 <+132>: mov r3, r7
0x0000883c <+136>: mov r0, #1
0x00008840 <+140>: ldr r1, [pc, #1292] ; 0x8d54 <main+1440>
0x00008844 <+144>: ldr r2, [pc, #1292] ; 0x8d58 <main+1444>
0x00008848 <+148>: bl 0x9ff0 <vgPlain_debugLog>
0x0000884c <+152>: cmp r4, #0
0x00008850 <+156>: beq 0x8bc8 <main+1044>
0x00008854 <+160>: mov r0, r4
0x00008858 <+164>: bl 0x8278 <select_platform>
0x0000885c <+168>: subs r8, r0, #0
0x00008860 <+172>: beq 0x8c6c <main+1208>
0x00008864 <+176>: mov r0, #1
0x00008868 <+180>: ldr r1, [pc, #1252] ; 0x8d54 <main+1440>
0x0000886c <+184>: ldr r2, [pc, #1256] ; 0x8d5c <main+1448>
0x00008870 <+188>: mov r3, r8
0x00008874 <+192>: bl 0x9ff0 <vgPlain_debugLog>
0x00008878 <+196>: mov r0, #0
0x0000887c <+200>: mov r5, r0
0x00008880 <+204>: add r4, r0, #500 ; 0x1f4
0x00008884 <+208>: mov r1, r4
0x00008888 <+212>: mov r0, r5
0x0000888c <+216>: bl 0x290ec <realloc>
0x00008890 <+220>: subs r5, r0, #0
0x00008894 <+224>: beq 0x8c88 <main+1236>
0x00008898 <+228>: ldr r0, [pc, #1216] ; 0x8d60 <main+1452>
0x0000889c <+232>: mov r1, r5
0x000088a0 <+236>: mov r2, r4
0x000088a4 <+240>: bl 0x2d760 <readlink>
0x000088a8 <+244>: cmn r0, #1
0x000088ac <+248>: beq 0x8c44 <main+1168>
0x000088b0 <+252>: cmp r0, r4
0x000088b4 <+256>: beq 0x8880 <main+204>
0x000088b8 <+260>: movcc r3, #0
0x000088bc <+264>: movcc r6, r5
0x000088c0 <+268>: strbcc r3, [r5, r0]
0x000088c4 <+272>: bcs 0x8d38 <main+1412>
0x000088c8 <+276>: ldr r0, [pc, #1172] ; 0x8d64 <main+1456>
0x000088cc <+280>: bl 0x29d20 <strlen>
0x000088d0 <+284>: mov r4, r0
0x000088d4 <+288>: mov r0, r6
0x000088d8 <+292>: bl 0x29d20 <strlen>
0x000088dc <+296>: add r4, r4, r0
0x000088e0 <+300>: add r0, r4, #2
0x000088e4 <+304>: bl 0x28928 <malloc>
0x000088e8 <+308>: subs r7, r0, #0
0x000088ec <+312>: beq 0x8ca4 <main+1264>
0x000088f0 <+316>: ldr r1, [pc, #1132] ; 0x8d64 <main+1456>
0x000088f4 <+320>: bl 0x29b00 <strcpy>
0x000088f8 <+324>: ldr r1, [pc, #1128] ; 0x8d68 <main+1460>
0x000088fc <+328>: mov r0, r7
0x00008900 <+332>: bl 0x29994 <strcat>
0x00008904 <+336>: mov r0, r7
0x00008908 <+340>: mov r1, r6
0x0000890c <+344>: bl 0x29994 <strcat>
0x00008910 <+348>: ldr r3, [r9]
0x00008914 <+352>: cmp r3, #0
0x00008918 <+356>: moveq r4, r3
0x0000891c <+360>: moveq r5, #2
0x00008920 <+364>: moveq r0, #8
0x00008924 <+368>: beq 0x8944 <main+400>
0x00008928 <+372>: mov r4, #0
0x0000892c <+376>: add r4, r4, #1
0x00008930 <+380>: ldr r3, [r9, r4, lsl #2]
0x00008934 <+384>: cmp r3, #0
0x00008938 <+388>: bne 0x892c <main+376>
0x0000893c <+392>: add r5, r4, #2
0x00008940 <+396>: lsl r0, r5, #2
0x00008944 <+400>: bl 0x28928 <malloc>
0x00008948 <+404>: subs r6, r0, #0
0x0000894c <+408>: beq 0x8cac <main+1272>
0x00008950 <+412>: cmp r4, #0
0x00008954 <+416>: moveq r1, r4
0x00008958 <+420>: moveq r0, #4
0x0000895c <+424>: moveq r2, #2
0x00008960 <+428>: beq 0x8994 <main+480>
0x00008964 <+432>: mov r1, #0
0x00008968 <+436>: mov r2, r1
0x0000896c <+440>: ldr r3, [r9, r2]
0x00008970 <+444>: add r1, r1, #1
0x00008974 <+448>: cmp r4, r1
0x00008978 <+452>: str r3, [r6, r2]
0x0000897c <+456>: add r2, r2, #4
0x00008980 <+460>: bgt 0x896c <main+440>
0x00008984 <+464>: add r3, r4, #1
0x00008988 <+468>: add r2, r3, #1
0x0000898c <+472>: lsl r1, r4, #2
0x00008990 <+476>: lsl r0, r3, #2
0x00008994 <+480>: cmp r2, r5
0x00008998 <+484>: str r7, [r6, r1]
0x0000899c <+488>: mov r3, #0
0x000089a0 <+492>: str r3, [r6, r0]
0x000089a4 <+496>: bne 0x8cb4 <main+1280>
0x000089a8 <+500>: ldr r0, [pc, #956] ; 0x8d6c <main+1464>
0x000089ac <+504>: bl 0x12650 <getenv>
0x000089b0 <+508>: subs r3, r0, #0
0x000089b4 <+512>: ldrne r0, [pc, #948] ; 0x8d70 <main+1468>
0x000089b8 <+516>: ldreq r0, [pc, #944] ; 0x8d70 <main+1468>
0x000089bc <+520>: strne r3, [r0]
0x000089c0 <+524>: ldr r0, [r0]
0x000089c4 <+528>: bl 0x29d20 <strlen>
0x000089c8 <+532>: mov r4, r0
0x000089cc <+536>: mov r0, r10
0x000089d0 <+540>: bl 0x29d20 <strlen>
0x000089d4 <+544>: mov r5, r0
0x000089d8 <+548>: mov r0, r8
0x000089dc <+552>: bl 0x29d20 <strlen>
0x000089e0 <+556>: add r4, r4, r5
0x000089e4 <+560>: add r4, r4, #3
0x000089e8 <+564>: add r0, r4, r0
0x000089ec <+568>: bl 0x28928 <malloc>
0x000089f0 <+572>: subs r4, r0, #0
0x000089f4 <+576>: bne 0x8cc8 <main+1300>
0x000089f8 <+580>: ldr r0, [pc, #884] ; 0x8d74 <main+1472>
0x000089fc <+584>: bl 0x8234 <barf>
0x00008a00 <+588>: ldrb r2, [r4, #1]
0x00008a04 <+592>: cmp r6, #1
0x00008a08 <+596>: movls r3, #0
0x00008a0c <+600>: movhi r3, #1
0x00008a10 <+604>: cmp r2, #45 ; 0x2d
0x00008a14 <+608>: movne r3, #0
0x00008a18 <+612>: cmp r3, #0
0x00008a1c <+616>: sub r0, r2, #45 ; 0x2d
0x00008a20 <+620>: beq 0x8a50 <main+668>
0x00008a24 <+624>: ldrb r0, [r4, #2]
0x00008a28 <+628>: cmp r6, #2
0x00008a2c <+632>: movls r3, #0
0x00008a30 <+636>: movhi r3, #1
0x00008a34 <+640>: cmp r0, #0
0x00008a38 <+644>: movne r3, #0
0x00008a3c <+648>: cmp r3, #0
0x00008a40 <+652>: ldrne r1, [pc, #772] ; 0x8d4c <main+1432>
0x00008a44 <+656>: ldrbne r2, [r4, #3]
0x00008a48 <+660>: ldrbne r3, [r1, #3]
0x00008a4c <+664>: rsbne r0, r3, r2
0x00008a50 <+668>: cmp r0, #0
0x00008a54 <+672>: beq 0x8810 <main+92>
0x00008a58 <+676>: ldr r0, [pc, #792] ; 0x8d78 <main+1476>
0x00008a5c <+680>: bl 0x29d20 <strlen>
0x00008a60 <+684>: cmp r0, #3
0x00008a64 <+688>: bhi 0x8c10 <main+1116>
0x00008a68 <+692>: cmp r0, #0
0x00008a6c <+696>: beq 0x8c08 <main+1108>
0x00008a70 <+700>: ldrb r2, [r4, #1]
0x00008a74 <+704>: cmp r0, #1
0x00008a78 <+708>: movls r3, #0
0x00008a7c <+712>: movhi r3, #1
0x00008a80 <+716>: cmp r2, #100 ; 0x64
0x00008a84 <+720>: movne r3, #0
0x00008a88 <+724>: cmp r3, #0
0x00008a8c <+728>: sub r2, r2, #100 ; 0x64
0x00008a90 <+732>: beq 0x8ac0 <main+780>
0x00008a94 <+736>: ldrb r2, [r4, #2]
0x00008a98 <+740>: cmp r0, #2
0x00008a9c <+744>: movls r3, #0
0x00008aa0 <+748>: movhi r3, #1
0x00008aa4 <+752>: cmp r2, #0
0x00008aa8 <+756>: movne r3, #0
0x00008aac <+760>: cmp r3, #0
0x00008ab0 <+764>: ldrne r1, [pc, #704] ; 0x8d78 <main+1476>
0x00008ab4 <+768>: ldrbne r2, [r4, #3]
0x00008ab8 <+772>: ldrbne r3, [r1, #3]
0x00008abc <+776>: rsbne r2, r3, r2
0x00008ac0 <+780>: cmp r2, #0
0x00008ac4 <+784>: beq 0x8c08 <main+1108>
0x00008ac8 <+788>: ldr r0, [pc, #684] ; 0x8d7c <main+1480>
0x00008acc <+792>: bl 0x29d20 <strlen>
0x00008ad0 <+796>: cmp r0, #6
0x00008ad4 <+800>: mov r1, r0
0x00008ad8 <+804>: bhi 0x8be4 <main+1072>
0x00008adc <+808>: cmp r0, #3
0x00008ae0 <+812>: bhi 0x8c24 <main+1136>
0x00008ae4 <+816>: cmp r0, #0
0x00008ae8 <+820>: beq 0x8b80 <main+972>
0x00008aec <+824>: ldrb r3, [r4, #1]
0x00008af0 <+828>: cmp r3, #45 ; 0x2d
0x00008af4 <+832>: movne r0, #0
0x00008af8 <+836>: moveq r0, #1
0x00008afc <+840>: cmp r1, #1
0x00008b00 <+844>: movls r2, #0
0x00008b04 <+848>: andhi r2, r0, #1
0x00008b08 <+852>: cmp r2, #0
0x00008b0c <+856>: beq 0x8b44 <main+912>
0x00008b10 <+860>: ldrb r3, [r4, #2]
0x00008b14 <+864>: cmp r3, #116 ; 0x74
0x00008b18 <+868>: movne r0, #0
0x00008b1c <+872>: moveq r0, #1
0x00008b20 <+876>: cmp r1, #2
0x00008b24 <+880>: movls r2, #0
0x00008b28 <+884>: andhi r2, r0, #1
0x00008b2c <+888>: cmp r2, #0
0x00008b30 <+892>: beq 0x8b44 <main+912>
0x00008b34 <+896>: ldrb r3, [r4, #3]
0x00008b38 <+900>: cmp r3, #111 ; 0x6f
0x00008b3c <+904>: movne r0, #0
0x00008b40 <+908>: moveq r0, #1
0x00008b44 <+912>: cmp r0, #0
0x00008b48 <+916>: bne 0x8b80 <main+972>
0x00008b4c <+920>: add r5, r5, #1
0x00008b50 <+924>: cmp r10, r5
0x00008b54 <+928>: ble 0x8b90 <main+988>
0x00008b58 <+932>: ldr r4, [r11, r5, lsl #2]
0x00008b5c <+936>: ldrb r3, [r4]
0x00008b60 <+940>: cmp r3, #45 ; 0x2d
0x00008b64 <+944>: bne 0x8820 <main+108>
0x00008b68 <+948>: cmp r6, #3
0x00008b6c <+952>: bls 0x8808 <main+84>
0x00008b70 <+956>: mov r0, r4
0x00008b74 <+960>: ldr r1, [pc, #464] ; 0x8d4c <main+1432>
0x00008b78 <+964>: bl 0x29acc <strcmp>
0x00008b7c <+968>: b 0x8a50 <main+668>
0x00008b80 <+972>: add r5, r5, #1
0x00008b84 <+976>: cmp r10, r5
0x00008b88 <+980>: add r7, r4, #7
0x00008b8c <+984>: bgt 0x8b58 <main+932>
0x00008b90 <+988>: mov r0, r8
0x00008b94 <+992>: ldr r1, [pc, #436] ; 0x8d50 <main+1436>
0x00008b98 <+996>: bl 0xa258 <vgPlain_debugLog_startup>
0x00008b9c <+1000>: cmp r7, #0
0x00008ba0 <+1004>: mov r10, r7
0x00008ba4 <+1008>: mov r4, #0
0x00008ba8 <+1012>: bne 0x8838 <main+132>
0x00008bac <+1016>: mov r0, #1
0x00008bb0 <+1020>: ldr r1, [pc, #412] ; 0x8d54 <main+1440>
0x00008bb4 <+1024>: ldr r2, [pc, #452] ; 0x8d80 <main+1484>
0x00008bb8 <+1028>: bl 0x9ff0 <vgPlain_debugLog>
0x00008bbc <+1032>: cmp r4, #0
0x00008bc0 <+1036>: ldr r10, [pc, #444] ; 0x8d84 <main+1488>
0x00008bc4 <+1040>: bne 0x8854 <main+160>
0x00008bc8 <+1044>: mov r0, #1
0x00008bcc <+1048>: ldr r1, [pc, #384] ; 0x8d54 <main+1440>
0x00008bd0 <+1052>: ldr r2, [pc, #432] ; 0x8d88 <main+1492>
0x00008bd4 <+1056>: ldr r3, [pc, #432] ; 0x8d8c <main+1496>
0x00008bd8 <+1060>: bl 0x9ff0 <vgPlain_debugLog>
0x00008bdc <+1064>: ldr r8, [pc, #424] ; 0x8d8c <main+1496>
0x00008be0 <+1068>: b 0x8878 <main+196>
0x00008be4 <+1072>: mov r0, r4
0x00008be8 <+1076>: ldr r1, [pc, #396] ; 0x8d7c <main+1480>
0x00008bec <+1080>: mov r2, #7
0x00008bf0 <+1084>: bl 0x29e90 <strncmp>
0x00008bf4 <+1088>: rsbs r0, r0, #1
0x00008bf8 <+1092>: movcc r0, #0
0x00008bfc <+1096>: cmp r0, #0
0x00008c00 <+1100>: beq 0x8b4c <main+920>
0x00008c04 <+1104>: b 0x8b80 <main+972>
0x00008c08 <+1108>: add r8, r8, #1
0x00008c0c <+1112>: b 0x8ac8 <main+788>
0x00008c10 <+1116>: mov r0, r4
0x00008c14 <+1120>: ldr r1, [pc, #348] ; 0x8d78 <main+1476>
0x00008c18 <+1124>: bl 0x29acc <strcmp>
0x00008c1c <+1128>: mov r2, r0
0x00008c20 <+1132>: b 0x8ac0 <main+780>
0x00008c24 <+1136>: mov r0, r4
0x00008c28 <+1140>: ldr r1, [pc, #332] ; 0x8d7c <main+1480>
0x00008c2c <+1144>: bl 0x29acc <strcmp>
0x00008c30 <+1148>: rsbs r0, r0, #1
0x00008c34 <+1152>: movcc r0, #0
0x00008c38 <+1156>: cmp r0, #0
0x00008c3c <+1160>: beq 0x8b4c <main+920>
0x00008c40 <+1164>: b 0x8b80 <main+972>
0x00008c44 <+1168>: ldr r4, [pc, #324] ; 0x8d90 <main+1500>
0x00008c48 <+1172>: ldr r1, [pc, #324] ; 0x8d94 <main+1504>
0x00008c4c <+1176>: ldr r2, [pc, #268] ; 0x8d60 <main+1452>
0x00008c50 <+1180>: ldr r0, [r4]
0x00008c54 <+1184>: bl 0x1b4c0 <fprintf>
0x00008c58 <+1188>: ldr r0, [r4]
0x00008c5c <+1192>: ldr r1, [pc, #308] ; 0x8d98 <main+1508>
0x00008c60 <+1196>: bl 0x1b4c0 <fprintf>
0x00008c64 <+1200>: ldr r6, [pc, #304] ; 0x8d9c <main+1512>
0x00008c68 <+1204>: b 0x88c8 <main+276>
0x00008c6c <+1208>: mov r0, #1
0x00008c70 <+1212>: ldr r1, [pc, #220] ; 0x8d54 <main+1440>
0x00008c74 <+1216>: ldr r2, [pc, #292] ; 0x8da0 <main+1516>
0x00008c78 <+1220>: ldr r3, [pc, #268] ; 0x8d8c <main+1496>
0x00008c7c <+1224>: bl 0x9ff0 <vgPlain_debugLog>
0x00008c80 <+1228>: ldr r8, [pc, #260] ; 0x8d8c <main+1496>
0x00008c84 <+1232>: b 0x8878 <main+196>
0x00008c88 <+1236>: ldr r0, [pc, #276] ; 0x8da4 <main+1520>
0x00008c8c <+1240>: bl 0x8234 <barf>
0x00008c90 <+1244>: mov r0, #0
0x00008c94 <+1248>: ldr r1, [pc, #180] ; 0x8d50 <main+1436>
0x00008c98 <+1252>: bl 0xa258 <vgPlain_debugLog_startup>
0x00008c9c <+1256>: mov r4, #0
0x00008ca0 <+1260>: b 0x8bac <main+1016>
0x00008ca4 <+1264>: ldr r0, [pc, #252] ; 0x8da8 <main+1524>
0x00008ca8 <+1268>: bl 0x8234 <barf>
0x00008cac <+1272>: ldr r0, [pc, #248] ; 0x8dac <main+1528>
0x00008cb0 <+1276>: bl 0x8234 <barf>
0x00008cb4 <+1280>: ldr r0, [pc, #244] ; 0x8db0 <main+1532>
0x00008cb8 <+1284>: ldr r1, [pc, #244] ; 0x8db4 <main+1536>
0x00008cbc <+1288>: movw r2, #463 ; 0x1cf
0x00008cc0 <+1292>: ldr r3, [pc, #240] ; 0x8db8 <main+1540>
0x00008cc4 <+1296>: bl 0xd5dc <__assert_fail>
0x00008cc8 <+1300>: ldr r3, [pc, #160] ; 0x8d70 <main+1468>
0x00008ccc <+1304>: ldr r1, [pc, #232] ; 0x8dbc <main+1544>
0x00008cd0 <+1308>: ldr r2, [r3]
0x00008cd4 <+1312>: mov r3, r10
0x00008cd8 <+1316>: str r8, [sp]
0x00008cdc <+1320>: bl 0x1b4e0 <sprintf>
0x00008ce0 <+1324>: mov r3, r4
0x00008ce4 <+1328>: mov r0, #1
0x00008ce8 <+1332>: ldr r1, [pc, #100] ; 0x8d54 <main+1440>
0x00008cec <+1336>: ldr r2, [pc, #204] ; 0x8dc0 <main+1548>
0x00008cf0 <+1340>: bl 0x9ff0 <vgPlain_debugLog>
0x00008cf4 <+1344>: mov r1, r11
0x00008cf8 <+1348>: mov r2, r6
0x00008cfc <+1352>: mov r0, r4
0x00008d00 <+1356>: bl 0x2ca60 <execve>
0x00008d04 <+1360>: ldr r3, [pc, #132] ; 0x8d90 <main+1500>
0x00008d08 <+1364>: ldr r4, [r3]
0x00008d0c <+1368>: bl 0xd5c0 <__errno_location>
0x00008d10 <+1372>: ldr r0, [r0]
0x00008d14 <+1376>: bl 0x29b60 <strerror>
0x00008d18 <+1380>: mov r2, r10
0x00008d1c <+1384>: mov r3, r8
0x00008d20 <+1388>: ldr r1, [pc, #156] ; 0x8dc4 <main+1552>
0x00008d24 <+1392>: str r0, [sp]
0x00008d28 <+1396>: mov r0, r4
0x00008d2c <+1400>: bl 0x1b4c0 <fprintf>
0x00008d30 <+1404>: mov r0, #1
0x00008d34 <+1408>: bl 0x12768 <exit>
0x00008d38 <+1412>: ldr r0, [pc, #136] ; 0x8dc8 <main+1556>
0x00008d3c <+1416>: ldr r1, [pc, #112] ; 0x8db4 <main+1536>
0x00008d40 <+1420>: movw r2, #438 ; 0x1b6
0x00008d44 <+1424>: ldr r3, [pc, #108] ; 0x8db8 <main+1540>
0x00008d48 <+1428>: bl 0xd5dc <__assert_fail>
0x00008d4c <+1432>: ; <UNDEFINED> instruction: 0x000672b4
0x00008d50 <+1436>: andeq r7, r6, r4, asr #5
0x00008d54 <+1440>: andeq r7, r6, r0, asr #2
0x00008d58 <+1444>: andeq r7, r6, r12, asr #5
0x00008d5c <+1448>: andeq r7, r6, r12, lsl #5
0x00008d60 <+1452>: andeq r7, r6, r0, lsr #7
0x00008d64 <+1456>: andeq r7, r6, r8, lsr r4
0x00008d68 <+1460>: andeq r8, r7, r0, asr #24
0x00008d6c <+1464>: muleq r6, r0, r4
0x00008d70 <+1468>: strdeq r4, [r8], -r0
0x00008d74 <+1472>: andeq r7, r6, r0, lsr #9
0x00008d78 <+1476>: ; <UNDEFINED> instruction: 0x000672b8
0x00008d7c <+1480>: ; <UNDEFINED> instruction: 0x000672bc
0x00008d80 <+1484>: andeq r7, r6, r4, ror #5
0x00008d84 <+1488>: andeq r7, r6, r4, lsl r3
0x00008d88 <+1492>: andeq r7, r6, r0, lsr #6
0x00008d8c <+1496>: andeq r7, r6, r8, lsl #4
0x00008d90 <+1500>: andeq r4, r8, r8, lsr #20
0x00008d94 <+1504>: ; <UNDEFINED> instruction: 0x000673b0
0x00008d98 <+1508>: andeq r7, r6, r8, ror #7
0x00008d9c <+1512>: andeq r9, r7, r8, lsl #29
0x00008da0 <+1516>: andeq r7, r6, r4, asr r3
0x00008da4 <+1520>: andeq r7, r6, r8, lsl #7
0x00008da8 <+1524>: andeq r7, r6, r12, asr #8
0x00008dac <+1528>: andeq r7, r6, r8, ror #8
0x00008db0 <+1532>: andeq r7, r6, r4, lsl #9
0x00008db4 <+1536>: muleq r6, r0, r1
0x00008db8 <+1540>: andeq r7, r6, r0, lsr r5
0x00008dbc <+1544>: ; <UNDEFINED> instruction: 0x000674bc
0x00008dc0 <+1548>: andeq r7, r6, r8, asr #9
0x00008dc4 <+1552>: ldrdeq r7, [r6], -r8
0x00008dc8 <+1556>: andeq r7, r6, r12, lsr #8

Related

Timer interrupts in ARM processor QEMU not occurring

I am trying to get interrupts to work on an ARM processor using QEMU. I have tried Google and various attempts to solve it, but I cannot get my IRQ handler to be called.
I also tried keyboard interrupts, which did not work either. I think the book I have may have a few mistakes, as the UART driver ended up being slightly different than described.
My assembly code:
.code 32
.global start, vectors_start, vectors_end, lock, unlock
start:
LDR sp, =stack_top
BL copy_vector
MSR cpsr, #0x92
LDR sp, =irq_stack_top
MSR cpsr, #0x13
bl unlock
BL main
B .
irq_handler:
sub lr, lr, #4
stmfd sp!, {r0-r12, lr}
bl IRQ_handler
ldmfd sp!, {r0-r12, pc}^
vectors_start:
LDR PC, reset_handler_addr
LDR PC, undef_handler_addr
LDR PC, swi_handler_addr
LDR PC, prefetch_abort_handler_addr
LDR PC, data_abort_handler_addr
B .
LDR PC, irq_handler_addr
LDR PC, fiq_handler_addr
reset_handler_addr: .word start
undef_handler_addr: .word undef_handler
swi_handler_addr: .word swi_handler
prefetch_abort_handler_addr: .word prefetch_abort_handler
data_abort_handler_addr: .word data_abort_handler
irq_handler_addr: .word irq_handler
fiq_handler_addr: .word fiq_handler
vectors_end:
lock:
MRS r0, CPSR
ORR r0,r0,#0x80
MSR CPSR, r0
mov PC, lr
unlock:
MRS r0,CPSR
BIC r0,r0,#0x80
MSR CPSR, r0
mov PC, lr
Relevant C functions:
void copy_vector(void){
extern u32 vectors_start, vectors_end;
u32 *vectors_src = &vectors_start;
u32 *vectors_dst = (u32 *) 0;
while (vectors_src<&vectors_end){
*vectors_dst++ = *vectors_src++;
}
}
// this function has been modified somewhat because the book was incomplete
// the print statement never prints
void IRQ_handler(void){
int vicstatus = 0;
uprints(uarts,"in interupt");
// int vicstatus = 0;
if(1){
if(*(tp[0]->base+TVALUE)==0){
timer_handler(0);
}
if(*(tp[1]->base+TVALUE)==0){
timer_handler(1);
}
}
if( 1){
if(*(tp[2]->base+TVALUE)==0){
timer_handler(2);
}
if(*(tp[3]->base+TVALUE)==0){
timer_handler(3);
}
}
}
void timer_init()
{
// timers[0].base = (u32*) 0x101E2000;
// timers[1].base = (u32*) 0x101E2020;
// timers[2].base = (u32*) 0x101E3000;
// timers[3].base = (u32*) 0x101E3020;
int i; TIMER * tp;
for (i = 0; i<4; ++i){
tp = &timers[i];
if(i==0) tp->base = (u32*) 0x101E2000;
if(i==1) tp->base = (u32*) 0x101E2020;
if (i == 2) tp->base = (u32*) 0x101E3000;
if (i == 3) tp->base = (u32*) 0x101E3020;
*(tp->base+TLOAD) = 0x0;
*(tp->base+TVALUE) = 0xFFFFFFFF;
*(tp->base+TRIS) = 0x0;
*(tp->base+TMIS) = 0x0;
*(tp->base+TLOAD) = 0x100;
*(tp->base+TCNTL) = 0x66;
*(tp->base+TBGLOAD) = 0x1C00;
*(tp->base ) |= (unsigned int)(1<<6);
tp->tick = 0;
///uprints(uarts,"timer init");
}
}
When I reset the machine in QEMU I get the message Timer with delta zero disabling.
I can get the message to disappear by changing: *(tp->base+TLOAD) = 0x0;
to *(tp->base+TLOAD) = 0x1; but the timer still doesn't work.
Lastly, my build script is:
arm-none-eabi-as -o start3.o start3.s
arm-none-eabi-gcc -c main3.c
arm-none-eabi-ld -T t.ld -o start3.elf start3.o main3.o
arm-none-eabi-nm start3.elf
arm-none-eabi-objcopy -O binary start3.elf start3.bin
qemu-system-arm -M versatilepb -m 128M -kernel start3.bin
edit: in addition here is the disassembled code from my main.c file (which includes a few other c files for the drivers)
the board is the versatilepb which according to the docs emulates the following:
ARM926E, ARM1136 or Cortex-A8 CPU
00000000 <uart_init>:
0: e52db004 push {fp} ; (str fp, [sp, #-4]!)
4: e28db000 add fp, sp, #0
8: e24dd00c sub sp, sp, #12
c: e3a03000 mov r3, #0
10: e50b3008 str r3, [fp, #-8]
14: ea000011 b 60 <uart_init+0x60>
18: e51b3008 ldr r3, [fp, #-8]
1c: e1a03183 lsl r3, r3, #3
20: e59f2064 ldr r2, [pc, #100] ; 8c <uart_init+0x8c>
24: e0833002 add r3, r3, r2
28: e50b300c str r3, [fp, #-12]
2c: e51b2008 ldr r2, [fp, #-8]
30: e59f3058 ldr r3, [pc, #88] ; 90 <uart_init+0x90>
34: e0823003 add r3, r2, r3
38: e1a03603 lsl r3, r3, #12
3c: e1a02003 mov r2, r3
40: e51b300c ldr r3, [fp, #-12]
44: e5832000 str r2, [r3]
48: e51b300c ldr r3, [fp, #-12]
4c: e51b2008 ldr r2, [fp, #-8]
50: e5832004 str r2, [r3, #4]
54: e51b3008 ldr r3, [fp, #-8]
58: e2833001 add r3, r3, #1
5c: e50b3008 str r3, [fp, #-8]
60: e51b3008 ldr r3, [fp, #-8]
64: e3530003 cmp r3, #3
68: daffffea ble 18 <uart_init+0x18>
6c: e59f3018 ldr r3, [pc, #24] ; 8c <uart_init+0x8c>
70: e59f201c ldr r2, [pc, #28] ; 94 <uart_init+0x94>
74: e5832018 str r2, [r3, #24]
78: e1a00000 nop ; (mov r0, r0)
7c: e1a00003 mov r0, r3
80: e28bd000 add sp, fp, #0
84: e49db004 pop {fp} ; (ldr fp, [sp], #4)
88: e12fff1e bx lr
8c: 00000000 .word 0x00000000
90: 000101f1 .word 0x000101f1
94: 10009000 .word 0x10009000
00000098 <ugetc>:
98: e52db004 push {fp} ; (str fp, [sp, #-4]!)
9c: e28db000 add fp, sp, #0
a0: e24dd00c sub sp, sp, #12
a4: e50b0008 str r0, [fp, #-8]
a8: e1a00000 nop ; (mov r0, r0)
ac: e51b3008 ldr r3, [fp, #-8]
b0: e5933000 ldr r3, [r3]
b4: e2833018 add r3, r3, #24
b8: e5d33000 ldrb r3, [r3]
bc: e2033010 and r3, r3, #16
c0: e3530000 cmp r3, #0
c4: 1afffff8 bne ac <ugetc+0x14>
c8: e51b3008 ldr r3, [fp, #-8]
cc: e5933000 ldr r3, [r3]
d0: e5d33000 ldrb r3, [r3]
d4: e1a00003 mov r0, r3
d8: e28bd000 add sp, fp, #0
dc: e49db004 pop {fp} ; (ldr fp, [sp], #4)
e0: e12fff1e bx lr
000000e4 <uputc>:
e4: e52db004 push {fp} ; (str fp, [sp, #-4]!)
e8: e28db000 add fp, sp, #0
ec: e24dd00c sub sp, sp, #12
f0: e50b0008 str r0, [fp, #-8]
f4: e1a03001 mov r3, r1
f8: e54b3009 strb r3, [fp, #-9]
fc: e1a00000 nop ; (mov r0, r0)
100: e51b3008 ldr r3, [fp, #-8]
104: e5933000 ldr r3, [r3]
108: e2833018 add r3, r3, #24
10c: e5d33000 ldrb r3, [r3]
110: e2033020 and r3, r3, #32
114: e3530000 cmp r3, #0
118: 1afffff8 bne 100 <uputc+0x1c>
11c: e51b3008 ldr r3, [fp, #-8]
120: e5933000 ldr r3, [r3]
124: e55b2009 ldrb r2, [fp, #-9]
128: e5c32000 strb r2, [r3]
12c: e1a00000 nop ; (mov r0, r0)
130: e1a00003 mov r0, r3
134: e28bd000 add sp, fp, #0
138: e49db004 pop {fp} ; (ldr fp, [sp], #4)
13c: e12fff1e bx lr
00000140 <ugets>:
140: e92d4800 push {fp, lr}
144: e28db004 add fp, sp, #4
148: e24dd008 sub sp, sp, #8
14c: e50b0008 str r0, [fp, #-8]
150: e50b100c str r1, [fp, #-12]
154: ea000007 b 178 <ugets+0x38>
158: e51b300c ldr r3, [fp, #-12]
15c: e5d33000 ldrb r3, [r3]
160: e1a01003 mov r1, r3
164: e51b0008 ldr r0, [fp, #-8]
168: ebfffffe bl e4 <uputc>
16c: e51b300c ldr r3, [fp, #-12]
170: e2833001 add r3, r3, #1
174: e50b300c str r3, [fp, #-12]
178: e51b0008 ldr r0, [fp, #-8]
17c: ebfffffe bl 98 <ugetc>
180: e1a03000 mov r3, r0
184: e20320ff and r2, r3, #255 ; 0xff
188: e51b300c ldr r3, [fp, #-12]
18c: e5c32000 strb r2, [r3]
190: e51b300c ldr r3, [fp, #-12]
194: e5d33000 ldrb r3, [r3]
198: e353000d cmp r3, #13
19c: 1affffed bne 158 <ugets+0x18>
1a0: e51b300c ldr r3, [fp, #-12]
1a4: e3a02000 mov r2, #0
1a8: e5c32000 strb r2, [r3]
1ac: e1a00000 nop ; (mov r0, r0)
1b0: e1a00003 mov r0, r3
1b4: e24bd004 sub sp, fp, #4
1b8: e8bd4800 pop {fp, lr}
1bc: e12fff1e bx lr
000001c0 <uprints>:
1c0: e92d4800 push {fp, lr}
1c4: e28db004 add fp, sp, #4
1c8: e24dd008 sub sp, sp, #8
1cc: e50b0008 str r0, [fp, #-8]
1d0: e50b100c str r1, [fp, #-12]
1d4: ea000006 b 1f4 <uprints+0x34>
1d8: e51b300c ldr r3, [fp, #-12]
1dc: e2832001 add r2, r3, #1
1e0: e50b200c str r2, [fp, #-12]
1e4: e5d33000 ldrb r3, [r3]
1e8: e1a01003 mov r1, r3
1ec: e51b0008 ldr r0, [fp, #-8]
1f0: ebfffffe bl e4 <uputc>
1f4: e51b300c ldr r3, [fp, #-12]
1f8: e5d33000 ldrb r3, [r3]
1fc: e3530000 cmp r3, #0
200: 1afffff4 bne 1d8 <uprints+0x18>
204: e1a00000 nop ; (mov r0, r0)
208: e1a00003 mov r0, r3
20c: e24bd004 sub sp, fp, #4
210: e8bd4800 pop {fp, lr}
214: e12fff1e bx lr
00000218 <fbuf_init>:
218: e52db004 push {fp} ; (str fp, [sp, #-4]!)
21c: e28db000 add fp, sp, #0
220: e59f3060 ldr r3, [pc, #96] ; 288 <fbuf_init+0x70>
224: e3a02602 mov r2, #2097152 ; 0x200000
228: e5832000 str r2, [r3]
22c: e59f3058 ldr r3, [pc, #88] ; 28c <fbuf_init+0x74>
230: e59f2058 ldr r2, [pc, #88] ; 290 <fbuf_init+0x78>
234: e5832000 str r2, [r3]
238: e59f3054 ldr r3, [pc, #84] ; 294 <fbuf_init+0x7c>
23c: e59f2054 ldr r2, [pc, #84] ; 298 <fbuf_init+0x80>
240: e5832000 str r2, [r3]
244: e59f3050 ldr r3, [pc, #80] ; 29c <fbuf_init+0x84>
248: e59f2050 ldr r2, [pc, #80] ; 2a0 <fbuf_init+0x88>
24c: e5832000 str r2, [r3]
250: e59f304c ldr r3, [pc, #76] ; 2a4 <fbuf_init+0x8c>
254: e59f204c ldr r2, [pc, #76] ; 2a8 <fbuf_init+0x90>
258: e5832000 str r2, [r3]
25c: e59f3048 ldr r3, [pc, #72] ; 2ac <fbuf_init+0x94>
260: e3a02602 mov r2, #2097152 ; 0x200000
264: e5832000 str r2, [r3]
268: e59f3040 ldr r3, [pc, #64] ; 2b0 <fbuf_init+0x98>
26c: e59f2040 ldr r2, [pc, #64] ; 2b4 <fbuf_init+0x9c>
270: e5832000 str r2, [r3]
274: e1a00000 nop ; (mov r0, r0)
278: e1a00003 mov r0, r3
27c: e28bd000 add sp, fp, #0
280: e49db004 pop {fp} ; (ldr fp, [sp], #4)
284: e12fff1e bx lr
288: 00000000 .word 0x00000000
28c: 1000001c .word 0x1000001c
290: 00002c77 .word 0x00002c77
294: 10120000 .word 0x10120000
298: 3f1f3f9c .word 0x3f1f3f9c
29c: 10120004 .word 0x10120004
2a0: 090b61df .word 0x090b61df
2a4: 10120008 .word 0x10120008
2a8: 071f1800 .word 0x071f1800
2ac: 10120010 .word 0x10120010
2b0: 10120018 .word 0x10120018
2b4: 0000082b .word 0x0000082b
000002b8 <draw_box>:
2b8: e52db004 push {fp} ; (str fp, [sp, #-4]!)
2bc: e28db000 add fp, sp, #0
2c0: e24dd01c sub sp, sp, #28
2c4: e50b0010 str r0, [fp, #-16]
2c8: e50b1014 str r1, [fp, #-20] ; 0xffffffec
2cc: e50b2018 str r2, [fp, #-24] ; 0xffffffe8
2d0: e50b301c str r3, [fp, #-28] ; 0xffffffe4
2d4: e3a03000 mov r3, #0
2d8: e50b3008 str r3, [fp, #-8]
2dc: ea00001c b 354 <draw_box+0x9c>
2e0: e3a03000 mov r3, #0
2e4: e50b300c str r3, [fp, #-12]
2e8: ea000012 b 338 <draw_box+0x80>
2ec: e59f3084 ldr r3, [pc, #132] ; 378 <draw_box+0xc0>
2f0: e5932000 ldr r2, [r3]
2f4: e51b100c ldr r1, [fp, #-12]
2f8: e51b3014 ldr r3, [fp, #-20] ; 0xffffffec
2fc: e0813003 add r3, r1, r3
300: e59f1074 ldr r1, [pc, #116] ; 37c <draw_box+0xc4>
304: e5911000 ldr r1, [r1]
308: e0010193 mul r1, r3, r1
30c: e51b3008 ldr r3, [fp, #-8]
310: e0811003 add r1, r1, r3
314: e51b3010 ldr r3, [fp, #-16]
318: e0813003 add r3, r1, r3
31c: e1a03103 lsl r3, r3, #2
320: e0823003 add r3, r2, r3
324: e51b201c ldr r2, [fp, #-28] ; 0xffffffe4
328: e5832000 str r2, [r3]
32c: e51b300c ldr r3, [fp, #-12]
330: e2833001 add r3, r3, #1
334: e50b300c str r3, [fp, #-12]
338: e51b200c ldr r2, [fp, #-12]
33c: e51b3018 ldr r3, [fp, #-24] ; 0xffffffe8
340: e1520003 cmp r2, r3
344: baffffe8 blt 2ec <draw_box+0x34>
348: e51b3008 ldr r3, [fp, #-8]
34c: e2833001 add r3, r3, #1
350: e50b3008 str r3, [fp, #-8]
354: e51b2008 ldr r2, [fp, #-8]
358: e51b3018 ldr r3, [fp, #-24] ; 0xffffffe8
35c: e1520003 cmp r2, r3
360: baffffde blt 2e0 <draw_box+0x28>
364: e1a00000 nop ; (mov r0, r0)
368: e1a00000 nop ; (mov r0, r0)
36c: e28bd000 add sp, fp, #0
370: e49db004 pop {fp} ; (ldr fp, [sp], #4)
374: e12fff1e bx lr
...
00000380 <timer_init>:
380: e52db004 push {fp} ; (str fp, [sp, #-4]!)
384: e28db000 add fp, sp, #0
388: e24dd00c sub sp, sp, #12
38c: e3a03000 mov r3, #0
390: e50b3008 str r3, [fp, #-8]
394: ea000048 b 4bc <timer_init+0x13c>
398: e51b3008 ldr r3, [fp, #-8]
39c: e1a03183 lsl r3, r3, #3
3a0: e59f2134 ldr r2, [pc, #308] ; 4dc <timer_init+0x15c>
3a4: e0833002 add r3, r3, r2
3a8: e50b300c str r3, [fp, #-12]
3ac: e51b3008 ldr r3, [fp, #-8]
3b0: e3530000 cmp r3, #0
3b4: 1a000002 bne 3c4 <timer_init+0x44>
3b8: e51b300c ldr r3, [fp, #-12]
3bc: e59f211c ldr r2, [pc, #284] ; 4e0 <timer_init+0x160>
3c0: e5832000 str r2, [r3]
3c4: e51b3008 ldr r3, [fp, #-8]
3c8: e3530001 cmp r3, #1
3cc: 1a000002 bne 3dc <timer_init+0x5c>
3d0: e51b300c ldr r3, [fp, #-12]
3d4: e59f2108 ldr r2, [pc, #264] ; 4e4 <timer_init+0x164>
3d8: e5832000 str r2, [r3]
3dc: e51b3008 ldr r3, [fp, #-8]
3e0: e3530002 cmp r3, #2
3e4: 1a000002 bne 3f4 <timer_init+0x74>
3e8: e51b300c ldr r3, [fp, #-12]
3ec: e59f20f4 ldr r2, [pc, #244] ; 4e8 <timer_init+0x168>
3f0: e5832000 str r2, [r3]
3f4: e51b3008 ldr r3, [fp, #-8]
3f8: e3530003 cmp r3, #3
3fc: 1a000002 bne 40c <timer_init+0x8c>
400: e51b300c ldr r3, [fp, #-12]
404: e59f20e0 ldr r2, [pc, #224] ; 4ec <timer_init+0x16c>
408: e5832000 str r2, [r3]
40c: e51b300c ldr r3, [fp, #-12]
410: e5933000 ldr r3, [r3]
414: e3a02000 mov r2, #0
418: e5832000 str r2, [r3]
41c: e51b300c ldr r3, [fp, #-12]
420: e5933000 ldr r3, [r3]
424: e2833004 add r3, r3, #4
428: e3e02000 mvn r2, #0
42c: e5832000 str r2, [r3]
430: e51b300c ldr r3, [fp, #-12]
434: e5933000 ldr r3, [r3]
438: e2833010 add r3, r3, #16
43c: e3a02000 mov r2, #0
440: e5832000 str r2, [r3]
444: e51b300c ldr r3, [fp, #-12]
448: e5933000 ldr r3, [r3]
44c: e2833014 add r3, r3, #20
450: e3a02000 mov r2, #0
454: e5832000 str r2, [r3]
458: e51b300c ldr r3, [fp, #-12]
45c: e5933000 ldr r3, [r3]
460: e3a02c01 mov r2, #256 ; 0x100
464: e5832000 str r2, [r3]
468: e51b300c ldr r3, [fp, #-12]
46c: e5933000 ldr r3, [r3]
470: e2833008 add r3, r3, #8
474: e3a02066 mov r2, #102 ; 0x66
478: e5832000 str r2, [r3]
47c: e51b300c ldr r3, [fp, #-12]
480: e5933000 ldr r3, [r3]
484: e2833018 add r3, r3, #24
488: e3a02b07 mov r2, #7168 ; 0x1c00
48c: e5832000 str r2, [r3]
490: e51b300c ldr r3, [fp, #-12]
494: e5933000 ldr r3, [r3]
498: e5932000 ldr r2, [r3]
49c: e3822040 orr r2, r2, #64 ; 0x40
4a0: e5832000 str r2, [r3]
4a4: e51b300c ldr r3, [fp, #-12]
4a8: e3a02000 mov r2, #0
4ac: e5832004 str r2, [r3, #4]
4b0: e51b3008 ldr r3, [fp, #-8]
4b4: e2833001 add r3, r3, #1
4b8: e50b3008 str r3, [fp, #-8]
4bc: e51b3008 ldr r3, [fp, #-8]
4c0: e3530003 cmp r3, #3
4c4: daffffb3 ble 398 <timer_init+0x18>
4c8: e1a00000 nop ; (mov r0, r0)
4cc: e1a00000 nop ; (mov r0, r0)
4d0: e28bd000 add sp, fp, #0
4d4: e49db004 pop {fp} ; (ldr fp, [sp], #4)
4d8: e12fff1e bx lr
4dc: 00000000 .word 0x00000000
4e0: 101e2000 .word 0x101e2000
4e4: 101e2020 .word 0x101e2020
4e8: 101e3000 .word 0x101e3000
4ec: 101e3020 .word 0x101e3020
000004f0 <timer_clearInterupt>:
4f0: e52db004 push {fp} ; (str fp, [sp, #-4]!)
4f4: e28db000 add fp, sp, #0
4f8: e24dd014 sub sp, sp, #20
4fc: e50b0010 str r0, [fp, #-16]
500: e51b3010 ldr r3, [fp, #-16]
504: e1a03183 lsl r3, r3, #3
508: e59f2028 ldr r2, [pc, #40] ; 538 <timer_clearInterupt+0x48>
50c: e0833002 add r3, r3, r2
510: e50b3008 str r3, [fp, #-8]
514: e51b3008 ldr r3, [fp, #-8]
518: e5933000 ldr r3, [r3]
51c: e283300c add r3, r3, #12
520: e3e02000 mvn r2, #0
524: e5832000 str r2, [r3]
528: e1a00000 nop ; (mov r0, r0)
52c: e28bd000 add sp, fp, #0
530: e49db004 pop {fp} ; (ldr fp, [sp], #4)
534: e12fff1e bx lr
538: 00000000 .word 0x00000000
0000053c <timer_handler>:
53c: e92d4800 push {fp, lr}
540: e28db004 add fp, sp, #4
544: e24dd010 sub sp, sp, #16
548: e50b0010 str r0, [fp, #-16]
54c: e51b3010 ldr r3, [fp, #-16]
550: e1a03183 lsl r3, r3, #3
554: e59f203c ldr r2, [pc, #60] ; 598 <timer_handler+0x5c>
558: e0833002 add r3, r3, r2
55c: e50b3008 str r3, [fp, #-8]
560: e51b3008 ldr r3, [fp, #-8]
564: e5933004 ldr r3, [r3, #4]
568: e2832001 add r2, r3, #1
56c: e51b3008 ldr r3, [fp, #-8]
570: e5832004 str r2, [r3, #4]
574: e51b0010 ldr r0, [fp, #-16]
578: ebfffffe bl 4f0 <timer_clearInterupt>
57c: e59f1018 ldr r1, [pc, #24] ; 59c <timer_handler+0x60>
580: e59f0018 ldr r0, [pc, #24] ; 5a0 <timer_handler+0x64>
584: ebfffffe bl 1c0 <uprints>
588: e1a00000 nop ; (mov r0, r0)
58c: e24bd004 sub sp, fp, #4
590: e8bd4800 pop {fp, lr}
594: e12fff1e bx lr
...
000005a4 <timer_start>:
5a4: e92d4800 push {fp, lr}
5a8: e28db004 add fp, sp, #4
5ac: e24dd010 sub sp, sp, #16
5b0: e50b0010 str r0, [fp, #-16]
5b4: e59f1044 ldr r1, [pc, #68] ; 600 <timer_start+0x5c>
5b8: e59f0044 ldr r0, [pc, #68] ; 604 <timer_start+0x60>
5bc: ebfffffe bl 1c0 <uprints>
5c0: e51b3010 ldr r3, [fp, #-16]
5c4: e1a03183 lsl r3, r3, #3
5c8: e59f2038 ldr r2, [pc, #56] ; 608 <timer_start+0x64>
5cc: e0833002 add r3, r3, r2
5d0: e50b3008 str r3, [fp, #-8]
5d4: e51b3008 ldr r3, [fp, #-8]
5d8: e5933000 ldr r3, [r3]
5dc: e2832008 add r2, r3, #8
5e0: e5922000 ldr r2, [r2]
5e4: e2833008 add r3, r3, #8
5e8: e3822080 orr r2, r2, #128 ; 0x80
5ec: e5832000 str r2, [r3]
5f0: e1a00000 nop ; (mov r0, r0)
5f4: e24bd004 sub sp, fp, #4
5f8: e8bd4800 pop {fp, lr}
5fc: e12fff1e bx lr
600: 00000010 .word 0x00000010
...
0000060c <timer_stop>:
60c: e52db004 push {fp} ; (str fp, [sp, #-4]!)
610: e28db000 add fp, sp, #0
614: e24dd014 sub sp, sp, #20
618: e50b0010 str r0, [fp, #-16]
61c: e51b3010 ldr r3, [fp, #-16]
620: e1a03183 lsl r3, r3, #3
624: e59f2028 ldr r2, [pc, #40] ; 654 <timer_stop+0x48>
628: e0833002 add r3, r3, r2
62c: e50b3008 str r3, [fp, #-8]
630: e51b3008 ldr r3, [fp, #-8]
634: e5933000 ldr r3, [r3]
638: e2833008 add r3, r3, #8
63c: e3a0207f mov r2, #127 ; 0x7f
640: e5832000 str r2, [r3]
644: e1a00000 nop ; (mov r0, r0)
648: e28bd000 add sp, fp, #0
64c: e49db004 pop {fp} ; (ldr fp, [sp], #4)
650: e12fff1e bx lr
654: 00000000 .word 0x00000000
00000658 <undef_handler>:
658: e92d4800 push {fp, lr}
65c: e28db004 add fp, sp, #4
660: e59f1014 ldr r1, [pc, #20] ; 67c <undef_handler+0x24>
664: e59f0014 ldr r0, [pc, #20] ; 680 <undef_handler+0x28>
668: ebfffffe bl 1c0 <uprints>
66c: e1a00000 nop ; (mov r0, r0)
670: e24bd004 sub sp, fp, #4
674: e8bd4800 pop {fp, lr}
678: e12fff1e bx lr
67c: 00000020 .word 0x00000020
680: 00000000 .word 0x00000000
00000684 <swi_handler>:
684: e92d4800 push {fp, lr}
688: e28db004 add fp, sp, #4
68c: e59f1014 ldr r1, [pc, #20] ; 6a8 <swi_handler+0x24>
690: e59f0014 ldr r0, [pc, #20] ; 6ac <swi_handler+0x28>
694: ebfffffe bl 1c0 <uprints>
698: e1a00000 nop ; (mov r0, r0)
69c: e24bd004 sub sp, fp, #4
6a0: e8bd4800 pop {fp, lr}
6a4: e12fff1e bx lr
6a8: 00000020 .word 0x00000020
6ac: 00000000 .word 0x00000000
000006b0 <prefetch_abort_handler>:
6b0: e92d4800 push {fp, lr}
6b4: e28db004 add fp, sp, #4
6b8: e59f1014 ldr r1, [pc, #20] ; 6d4 <prefetch_abort_handler+0x24>
6bc: e59f0014 ldr r0, [pc, #20] ; 6d8 <prefetch_abort_handler+0x28>
6c0: ebfffffe bl 1c0 <uprints>
6c4: e1a00000 nop ; (mov r0, r0)
6c8: e24bd004 sub sp, fp, #4
6cc: e8bd4800 pop {fp, lr}
6d0: e12fff1e bx lr
6d4: 00000020 .word 0x00000020
6d8: 00000000 .word 0x00000000
000006dc <data_abort_handler>:
6dc: e92d4800 push {fp, lr}
6e0: e28db004 add fp, sp, #4
6e4: e59f1014 ldr r1, [pc, #20] ; 700 <data_abort_handler+0x24>
6e8: e59f0014 ldr r0, [pc, #20] ; 704 <data_abort_handler+0x28>
6ec: ebfffffe bl 1c0 <uprints>
6f0: e1a00000 nop ; (mov r0, r0)
6f4: e24bd004 sub sp, fp, #4
6f8: e8bd4800 pop {fp, lr}
6fc: e12fff1e bx lr
700: 00000020 .word 0x00000020
704: 00000000 .word 0x00000000
00000708 <fiq_handler>:
708: e92d4800 push {fp, lr}
70c: e28db004 add fp, sp, #4
710: e59f1014 ldr r1, [pc, #20] ; 72c <fiq_handler+0x24>
714: e59f0014 ldr r0, [pc, #20] ; 730 <fiq_handler+0x28>
718: ebfffffe bl 1c0 <uprints>
71c: e1a00000 nop ; (mov r0, r0)
720: e24bd004 sub sp, fp, #4
724: e8bd4800 pop {fp, lr}
728: e12fff1e bx lr
72c: 00000020 .word 0x00000020
730: 00000000 .word 0x00000000
00000734 <copy_vector>:
734: e52db004 push {fp} ; (str fp, [sp, #-4]!)
738: e28db000 add fp, sp, #0
73c: e24dd00c sub sp, sp, #12
740: e59f3050 ldr r3, [pc, #80] ; 798 <copy_vector+0x64>
744: e50b3008 str r3, [fp, #-8]
748: e3a03000 mov r3, #0
74c: e50b300c str r3, [fp, #-12]
750: ea000007 b 774 <copy_vector+0x40>
754: e51b2008 ldr r2, [fp, #-8]
758: e2823004 add r3, r2, #4
75c: e50b3008 str r3, [fp, #-8]
760: e51b300c ldr r3, [fp, #-12]
764: e2831004 add r1, r3, #4
768: e50b100c str r1, [fp, #-12]
76c: e5922000 ldr r2, [r2]
770: e5832000 str r2, [r3]
774: e51b3008 ldr r3, [fp, #-8]
778: e59f201c ldr r2, [pc, #28] ; 79c <copy_vector+0x68>
77c: e1530002 cmp r3, r2
780: 3afffff3 bcc 754 <copy_vector+0x20>
784: e1a00000 nop ; (mov r0, r0)
788: e1a00000 nop ; (mov r0, r0)
78c: e28bd000 add sp, fp, #0
790: e49db004 pop {fp} ; (ldr fp, [sp], #4)
794: e12fff1e bx lr
...
000007a0 <kbd_init>:
7a0: e52db004 push {fp} ; (str fp, [sp, #-4]!)
7a4: e28db000 add fp, sp, #0
7a8: e24dd00c sub sp, sp, #12
7ac: e59f302c ldr r3, [pc, #44] ; 7e0 <kbd_init+0x40>
7b0: e50b3008 str r3, [fp, #-8]
7b4: e51b3008 ldr r3, [fp, #-8]
7b8: e3a02014 mov r2, #20
7bc: e5c32000 strb r2, [r3]
7c0: e51b3008 ldr r3, [fp, #-8]
7c4: e283300c add r3, r3, #12
7c8: e3a02008 mov r2, #8
7cc: e5c32000 strb r2, [r3]
7d0: e1a00000 nop ; (mov r0, r0)
7d4: e28bd000 add sp, fp, #0
7d8: e49db004 pop {fp} ; (ldr fp, [sp], #4)
7dc: e12fff1e bx lr
7e0: 10006000 .word 0x10006000
000007e4 <IRQ_handler>:
7e4: e92d4800 push {fp, lr}
7e8: e28db004 add fp, sp, #4
7ec: e24dd008 sub sp, sp, #8
7f0: e3a03000 mov r3, #0
7f4: e50b3008 str r3, [fp, #-8]
7f8: e59f10a4 ldr r1, [pc, #164] ; 8a4 <IRQ_handler+0xc0>
7fc: e59f00a4 ldr r0, [pc, #164] ; 8a8 <IRQ_handler+0xc4>
800: ebfffffe bl 1c0 <uprints>
804: e59f30a0 ldr r3, [pc, #160] ; 8ac <IRQ_handler+0xc8>
808: e5933000 ldr r3, [r3]
80c: e5933000 ldr r3, [r3]
810: e2833004 add r3, r3, #4
814: e5933000 ldr r3, [r3]
818: e3530000 cmp r3, #0
81c: 1a000001 bne 828 <IRQ_handler+0x44>
820: e3a00000 mov r0, #0
824: ebfffffe bl 53c <timer_handler>
828: e59f307c ldr r3, [pc, #124] ; 8ac <IRQ_handler+0xc8>
82c: e5933004 ldr r3, [r3, #4]
830: e5933000 ldr r3, [r3]
834: e2833004 add r3, r3, #4
838: e5933000 ldr r3, [r3]
83c: e3530000 cmp r3, #0
840: 1a000001 bne 84c <IRQ_handler+0x68>
844: e3a00001 mov r0, #1
848: ebfffffe bl 53c <timer_handler>
84c: e59f3058 ldr r3, [pc, #88] ; 8ac <IRQ_handler+0xc8>
850: e5933008 ldr r3, [r3, #8]
854: e5933000 ldr r3, [r3]
858: e2833004 add r3, r3, #4
85c: e5933000 ldr r3, [r3]
860: e3530000 cmp r3, #0
864: 1a000001 bne 870 <IRQ_handler+0x8c>
868: e3a00002 mov r0, #2
86c: ebfffffe bl 53c <timer_handler>
870: e59f3034 ldr r3, [pc, #52] ; 8ac <IRQ_handler+0xc8>
874: e593300c ldr r3, [r3, #12]
878: e5933000 ldr r3, [r3]
87c: e2833004 add r3, r3, #4
880: e5933000 ldr r3, [r3]
884: e3530000 cmp r3, #0
888: 1a000001 bne 894 <IRQ_handler+0xb0>
88c: e3a00003 mov r0, #3
890: ebfffffe bl 53c <timer_handler>
894: e1a00000 nop ; (mov r0, r0)
898: e24bd004 sub sp, fp, #4
89c: e8bd4800 pop {fp, lr}
8a0: e12fff1e bx lr
8a4: 00000030 .word 0x00000030
...
000008b0 <main>:
8b0: e92d4800 push {fp, lr}
8b4: e28db004 add fp, sp, #4
8b8: e24dd010 sub sp, sp, #16
8bc: ebfffffe bl 218 <fbuf_init>
8c0: ebfffffe bl 0 <uart_init>
8c4: ebfffffe bl 380 <timer_init>
8c8: ebfffffe bl 7a0 <kbd_init>
8cc: e3a03000 mov r3, #0
8d0: e50b3008 str r3, [fp, #-8]
8d4: ea00000b b 908 <main+0x58>
8d8: e51b3008 ldr r3, [fp, #-8]
8dc: e1a03183 lsl r3, r3, #3
8e0: e59f20dc ldr r2, [pc, #220] ; 9c4 <main+0x114>
8e4: e0832002 add r2, r3, r2
8e8: e59f10d8 ldr r1, [pc, #216] ; 9c8 <main+0x118>
8ec: e51b3008 ldr r3, [fp, #-8]
8f0: e7812103 str r2, [r1, r3, lsl #2]
8f4: e51b0008 ldr r0, [fp, #-8]
8f8: ebfffffe bl 5a4 <timer_start>
8fc: e51b3008 ldr r3, [fp, #-8]
900: e2833001 add r3, r3, #1
904: e50b3008 str r3, [fp, #-8]
908: e51b3008 ldr r3, [fp, #-8]
90c: e3530003 cmp r3, #3
910: dafffff0 ble 8d8 <main+0x28>
914: e59f10b0 ldr r1, [pc, #176] ; 9cc <main+0x11c>
918: e59f00b0 ldr r0, [pc, #176] ; 9d0 <main+0x120>
91c: ebfffffe bl 1c0 <uprints>
920: e59f309c ldr r3, [pc, #156] ; 9c4 <main+0x114>
924: e5931004 ldr r1, [r3, #4]
928: e59f30a4 ldr r3, [pc, #164] ; 9d4 <main+0x124>
92c: e0c32391 smull r2, r3, r1, r3
930: e1a021c3 asr r2, r3, #3
934: e1a03fc1 asr r3, r1, #31
938: e0422003 sub r2, r2, r3
93c: e1a03002 mov r3, r2
940: e1a03183 lsl r3, r3, #3
944: e0433002 sub r3, r3, r2
948: e1a03183 lsl r3, r3, #3
94c: e0433002 sub r3, r3, r2
950: e1a03103 lsl r3, r3, #2
954: e0412003 sub r2, r1, r3
958: e1a03002 mov r3, r2
95c: e54b3009 strb r3, [fp, #-9]
960: e55b3009 ldrb r3, [fp, #-9]
964: e2833037 add r3, r3, #55 ; 0x37
968: e20330ff and r3, r3, #255 ; 0xff
96c: e1a01003 mov r1, r3
970: e59f0058 ldr r0, [pc, #88] ; 9d0 <main+0x120>
974: ebfffffe bl e4 <uputc>
978: e55b3009 ldrb r3, [fp, #-9]
97c: e383386e orr r3, r3, #7208960 ; 0x6e0000
980: e3833001 orr r3, r3, #1
984: e50b3010 str r3, [fp, #-16]
988: e51b3010 ldr r3, [fp, #-16]
98c: e3a02064 mov r2, #100 ; 0x64
990: e3a0100a mov r1, #10
994: e3a0000a mov r0, #10
998: ebfffffe bl 2b8 <draw_box>
99c: e59f3020 ldr r3, [pc, #32] ; 9c4 <main+0x114>
9a0: e5933004 ldr r3, [r3, #4]
9a4: e3530000 cmp r3, #0
9a8: 0affffdc beq 920 <main+0x70>
9ac: e3a03b32 mov r3, #51200 ; 0xc800
9b0: e3a02032 mov r2, #50 ; 0x32
9b4: e3a010fa mov r1, #250 ; 0xfa
9b8: e3a000c8 mov r0, #200 ; 0xc8
9bc: ebfffffe bl 2b8 <draw_box>
9c0: eaffffd6 b 920 <main+0x70>
...
9cc: 0000003c .word 0x0000003c
9d0: 00000000 .word 0x00000000
9d4: 094f2095 .word 0x094f2095
also my linker script:
ENTRY(start)
SECTIONS{
. = 0x10000;
.text : {*(.text)}
.data : {*(.data)}
.bss : {*(.bss)}
. =ALIGN(8);
. =. + 0x1000;
stack_top = .;
. = . + 0x1000;
irq_stack_top = .;
}
So I do not have a complete example but something looks wrong, really wrong that you will need to sort out. And maybe you are already through all of these basic steps to get to this point.
qemu-system-arm -M versatilepb -m 128M -kernel notmain.bin
Implies a binary memory image file as in objcopy -O binary is being used. And there is no addressing information, so some default built into that machine or simulation is being used. What is that address....
strap.s
mov sp,#0x10000
bl notmain
b .
.globl GETPC
GETPC:
mov r0,pc
bx lr
.globl PUT32
PUT32:
str r1,[r0]
bx lr
notmain.c
void PUT32 ( unsigned int, unsigned int );
unsigned int GETPC ( void );
#define UART_BASE 0x101F1000
#define UARTDR (UART_BASE+0x000)
static void uart_send ( unsigned int x )
{
PUT32(UARTDR,x);
}
static void hexstrings ( unsigned int d )
{
unsigned int rb;
unsigned int rc;
rb=32;
while(1)
{
rb-=4;
rc=(d>>rb)&0xF;
if(rc>9) rc+=0x37; else rc+=0x30;
uart_send(rc);
if(rb==0) break;
}
uart_send(0x20);
}
static void hexstring ( unsigned int d )
{
hexstrings(d);
uart_send(0x0D);
uart_send(0x0A);
}
int notmain ( void )
{
hexstring(GETPC());
return(0);
}
uart address culled from the source code to qemu. so far I find you can cheat and just write to the tx buffer, it does not actually emulate the time it takes to send out a character and overflow a fifo, etc. So I cheat.
/* memmap */
MEMORY
{
ram : ORIGIN = 0x00000000, LENGTH = 32K
}
SECTIONS
{
.text : { *(.text*) } > ram
.bss : { *(.text*) } > ram
}
build
arm-none-eabi-as --warn --fatal-warnings -march=armv4t strap.s -o strap.o
arm-none-eabi-gcc -c -Wall -O2 -ffreestanding -march=armv4t notmain.c -o notmain.o
arm-none-eabi-ld -nostdlib -nostartfiles strap.o notmain.o -T memmap -o notmain.elf
arm-none-eabi-objdump -D notmain.elf > notmain.list
arm-none-eabi-objcopy notmain.elf -O binary notmain.bin
examine output
Disassembly of section .text:
00000000 <GETPC-0xc>:
0: e3a0d801 mov sp, #65536 ; 0x10000
4: eb000004 bl 1c <notmain>
8: eafffffe b 8 <GETPC-0x4>
0000000c <GETPC>:
c: e1a0000f mov r0, pc
10: e12fff1e bx lr
00000014 <PUT32>:
14: e5801000 str r1, [r0]
18: e12fff1e bx lr
0000001c <notmain>:
1c: e92d4070 push {r4, r5, r6, lr}
20: ebfffff9 bl c <GETPC>
24: e3a04020 mov r4, #32
my entry point is where I want it. I am not loading an elf I am loading a binary image. If I wanted the entry point to be different and load an elf it may be possible that using ENTRY() in the linker script then having the entry point wherever might work.
qemu-system-arm -M versatilepb -m 128M -kernel notmain.bin
then view the serial0 output of 00010014 so they are loading us at 0x10000 so change memmap
ram : ORIGIN = 0x00010000, LENGTH = 32K
Now your linker script implies it should have put your code at 0x10000 but
clearly it has not based on your disassembly, so how are you running anything at all. Also your entry point using a binary image is not the bootstrap it is not the first instructions you want to be executing. maybe you disassembled an object not the final binary?
So for fun before changing the memory map to 0x10000 if I instead
qemu-system-arm -M versatilepb -m 128M -kernel notmain.elf
Then I get 00000014 on the output. it used the address defined in the elf file.
If I add this
b .
.globl _start
_start:
mov sp,#0x10000
bl notmain
b .
and this
/* memmap */
ENTRY(_start)
MEMORY
{
ram : ORIGIN = 0x00000000, LENGTH = 32K
}
SECTIONS
{
.text : { *(.text*) } > ram
.bss : { *(.text*) } > ram
}
and execute the elf I do get output of 00000018 so it is entering at the entry point not the first word of the loadable image.
run the .bin and you get no serial output as expected.
Change memmap to 0x10000 and run the elf you get 0x10018 as expected.
So that was a simple way to see how qemu loads our program.
so same build and linker script (with the 0x10000)
.globl _start
_start:
ldr pc,reset_handler
ldr pc,undefined_handler
ldr pc,swi_handler
ldr pc,prefetch_handler
ldr pc,data_handler
ldr pc,unused_handler
ldr pc,irq_handler
ldr pc,fiq_handler
reset_handler: .word reset
undefined_handler: .word hang
swi_handler: .word swi
prefetch_handler: .word hang
data_handler: .word hang
unused_handler: .word hang
irq_handler: .word hang
fiq_handler: .word hang
reset:
mov r0,#0x10000
mov r1,#0x00000
ldmia r0!,{r2,r3,r4,r5,r6,r7,r8,r9}
stmia r1!,{r2,r3,r4,r5,r6,r7,r8,r9}
ldmia r0!,{r2,r3,r4,r5,r6,r7,r8,r9}
stmia r1!,{r2,r3,r4,r5,r6,r7,r8,r9}
;# FIQ 110 10001
msr cpsr_c,#0xD1
mov sp,#0x1000
;# IRQ 110 10010
msr cpsr_c,#0xD2
mov sp,#0x2000
;# SVC 110 10011
msr cpsr_c,#0xD3
mov sp,#0x3000
;# ABT 110 10111
msr cpsr_c,#0xD7
mov sp,#0x4000
;# UND 110 11011
msr cpsr_c,#0xDB
mov sp,#0x5000
;# SYS 110 11111
msr cpsr_c,#0xDF
mov sp,#0x6000
;# SVC 110 10011
msr cpsr_c,#0xD3
bl notmain
hang:
b hang
swi:
push {r4-r12,lr}
bl swi_main
pop {r4-r12,lr}
movs pc,lr
.globl DO_SWI
DO_SWI:
push {lr}
svc 0
pop {lr}
bx lr
.globl GO_USER
GO_USER:
;# USER 110 10000
msr cpsr_c,#0xD0
bl gouser
b .
gouser:
bx r0
.globl GETMODE
GETMODE:
mrs r0,cpsr
bx lr
.globl GETPC
GETPC:
mov r0,pc
bx lr
.globl GETSP
GETSP:
mov r0,sp
bx lr
.globl PUT32
PUT32:
str r1,[r0]
bx lr
.globl GET32
GET32:
ldr r0,[r0]
bx lr
and
void PUT32 ( unsigned int, unsigned int );
unsigned int GET32 ( unsigned int );
unsigned int GETPC ( void );
unsigned int GETSP ( void );
unsigned int GETMODE ( void );
unsigned int GO_USER ( unsigned int );
unsigned int DO_SWI ( unsigned int, unsigned int, unsigned int, unsigned int );
#define UART_BASE 0x101F1000
#define UARTDR (UART_BASE+0x000)
#define UARTFR (UART_BASE+0x018)
static void uart_send ( unsigned int x )
{
while(1)
{
if(GET32(UARTFR)&(1<<7)) break;
}
PUT32(UARTDR,x);
}
static void hexstrings ( unsigned int d )
{
unsigned int rb;
unsigned int rc;
rb=32;
while(1)
{
rb-=4;
rc=(d>>rb)&0xF;
if(rc>9) rc+=0x37; else rc+=0x30;
uart_send(rc);
if(rb==0) break;
}
uart_send(0x20);
}
static void hexstring ( unsigned int d )
{
hexstrings(d);
uart_send(0x0D);
uart_send(0x0A);
}
unsigned int swi_main ( unsigned int a, unsigned int b, unsigned int c, unsigned int d )
{
hexstring(0xAABBCCDD);
switch(a)
{
case 0:
{
return(~b);
}
case 1:
{
return(GETMODE());
}
}
return(0);
}
void user_main ( void )
{
hexstring(GETMODE());
hexstring(GETSP());
hexstring(DO_SWI(0,0x1234,0,0));
hexstring(DO_SWI(1,0x1234,0,0));
}
int notmain ( void )
{
hexstring(GETMODE());
hexstring(GETSP());
GO_USER((unsigned int)user_main);
return(0);
}
get even a trivial swi in place first before attempting an interrupt. getting the exception table in place, the banked stack registers, etc. seems trivial but even with experience those things bite you and you end up chasing the wrong thing.
400001D3
00002FF8
200001D0
00005FF8
AABBCCDD
FFFFEDCB
AABBCCDD
600001D3
this was a bit overkill with mode changes and such, but it still has the swi handler in the right place. The code is linked for
notmain.elf: file format elf32-littlearm
Disassembly of section .text:
00010000 <_start>:
10000: e59ff018 ldr pc, [pc, #24] ; 10020 <reset_handler>
10004: e59ff018 ldr pc, [pc, #24] ; 10024 <undefined_handler>
10008: e59ff018 ldr pc, [pc, #24] ; 10028 <swi_handler>
1000c: e59ff018 ldr pc, [pc, #24] ; 1002c <prefetch_handler>
10010: e59ff018 ldr pc, [pc, #24] ; 10030 <data_handler>
10014: e59ff018 ldr pc, [pc, #24] ; 10034 <unused_handler>
10018: e59ff018 ldr pc, [pc, #24] ; 10038 <irq_handler>
1001c: e59ff018 ldr pc, [pc, #24] ; 1003c <fiq_handler>
00010020 <reset_handler>:
10020: 00010040 andeq r0, r1, r0, asr #32
00010024 <undefined_handler>:
10024: 00010090 muleq r1, r0, r0
packing the exception table with its pc-relative addresses and letting the toolchain do all the work of sorting out addresses for you, you can then simply copy it to 0x00000000 from 0x10000. It is probably fine but in your case I would at least put some level of alignment before your table, something makes me want to think it wants to be aligned. However you wish to copy it is your deal, I would not use C in a bootstrap, the habit creates problems from time to time that are easily avoided.
Knowing what we know of course if we linked for address 0x00000000 and used the elf file then we do not have to copy/move the exception table. It lands in the right place on load.
Newer cores have a VTOR register and you can simply choose a different exception or vector table address, but it does need to have a level of alignment that you cannot forget.
I did not go and research this timer, but I assume long before you attempted an interrupt you wrote very many throwaway programs that used polling to fully understand the timer and how to clear its interrupt, etc? Depending on the chip (not necessarily core) then there may also be arm side registers you can poll, disable interrupts into the core, poll the gvic or vpic or whatever it is called and see the interrupt hit the edge of the core, be able to turn it off, etc. Then.....enable it into the core now that you have mastered that peripheral. Sometimes leaving nothing left to work on, for some cores though there is some interrupt stuff that can only be done in an isr (how to detect and clear the interrupt source) this virtual machine predates that by a long time though, so no worries there.
so...
If that is your linker script then why is your output zero based, and why is the uart function up front and not the bootstrap entry point? If that is the disassembly of the final binary then there are all kinds of problems going on there.
I was seeing this as baremetal but forgetting qemu is loading linux-like so at some address like 0x8000, 0x10000, or 0x80000 are some examples, so yes you either need a core with a VTOR or copy the exception table over, and letting the tools build the addresses and pc-relative loads is the way to go. That copy is probably fine if you are actually booting right and executing it.
Then have you completely mastered that peripheral?
This is an emulated core so it may resemble one/some of the products from arm but remember they are not all identical some come with interrupt controllers some do not, some folks like the raspi folks disable that interrupt controller, etc. The base address of add on peripherals for real arm cores is based on the chip vendor not arm. Etc....Best to dig through the source code for the specific machine emulation and see what it is doing to truly get a qemu machine to work.
So maybe you have done all of these things and it still does not work. The question as written has some issues, the disassembly does not make sense for example. Need to resolve that. What if you try with an elf file first and an entry point then do the extra work to be able to run without the elf file and the bin file instead. Do not re-write but add more content to the question if you need further help. Unless someone else sees the issue with what has been presented.

Linking in .rodata and .debug_frame using arm-none-eabi-gcc

I am trying to link my .o files to generate the ELF to be run on an ARM M0+ bare metal. I am using arm-none-eabi-gcc. If I just have my C file (s) written it links fine and I can run it no problem. (For this I am providing the -lnosys -lc -lm flags to GCC)
However for my purpose I need to 'overwrite' one of the library functions, namely __aeabi_fmul. To do this, I took the disassembly of this function and created a custom file aeabi_fmul.S which I compile separately. Now I need to link this aeabi_fmul.o file with my C source code (so that the final binary will use my implementation of fmul instead of the library's)
It seems to be linking this file in fine (the listing shows that the binary content is identical here), however, if I do an objdump -D there doesn't seem to be any .rodata or ˚ sections in my 'modified' ELF.
This is causing my program to 'randomly' hang while executing a branch instruction, which I'm guessing is linked to these missing sections (because everything else matches 1:1 with the 'working' version)
Is there something else I need to do to get these two sections to be linked?
Could the fact that I am linking in my own __aeabi_fmul be causing these sections to not get linked?
EDIT:
Dissassembly of working code.
main.elf: file format elf32-littlearm
Disassembly of section .text:
00000000 <hang-0x50>:
0: 40800000 addmi r0, r0, r0
4: 00000053 andeq r0, r0, r3, asr r0
8: 00000051 andeq r0, r0, r1, asr r0
c: 00000051 andeq r0, r0, r1, asr r0
10: 00000051 andeq r0, r0, r1, asr r0
14: 00000051 andeq r0, r0, r1, asr r0
18: 00000051 andeq r0, r0, r1, asr r0
1c: 00000051 andeq r0, r0, r1, asr r0
20: 00000051 andeq r0, r0, r1, asr r0
24: 00000051 andeq r0, r0, r1, asr r0
28: 00000051 andeq r0, r0, r1, asr r0
2c: 00000051 andeq r0, r0, r1, asr r0
30: 00000051 andeq r0, r0, r1, asr r0
34: 00000051 andeq r0, r0, r1, asr r0
38: 00000051 andeq r0, r0, r1, asr r0
3c: 00000051 andeq r0, r0, r1, asr r0
40: 00000051 andeq r0, r0, r1, asr r0
44: 00000051 andeq r0, r0, r1, asr r0
48: 00000051 andeq r0, r0, r1, asr r0
4c: 00000051 andeq r0, r0, r1, asr r0
00000050 <hang>:
50: e7fe b.n 50 <hang>
00000052 <_start>:
52: f000 f89b bl 18c <main>
00000056 <exit>:
56: df01 svc 1
58: e7fe b.n 58 <exit+0x2>
0000005a <putchar>:
5a: b580 push {r7, lr}
5c: af00 add r7, sp, #0
5e: 22e0 movs r2, #224 ; 0xe0
60: 0612 lsls r2, r2, #24
62: 6010 str r0, [r2, #0]
64: 46bd mov sp, r7
66: bd80 pop {r7, pc}
00000068 <PUT32>:
68: 6001 str r1, [r0, #0]
6a: 4770 bx lr
0000006c <GET32>:
6c: 6800 ldr r0, [r0, #0]
6e: 4770 bx lr
00000070 <PUTGETCLR>:
70: 6802 ldr r2, [r0, #0]
72: 438a bics r2, r1
74: 6002 str r2, [r0, #0]
76: 4770 bx lr
00000078 <atest>:
78: 1842 adds r2, r0, r1
7a: 2000 movs r0, #0
7c: d701 bvc.n 82 <atest1>
7e: 2301 movs r3, #1
80: 4318 orrs r0, r3
00000082 <atest1>:
82: d301 bcc.n 88 <atest2>
84: 2302 movs r3, #2
86: 4318 orrs r0, r3
00000088 <atest2>:
88: 4770 bx lr
0000008a <stest>:
8a: 1a42 subs r2, r0, r1
8c: 2000 movs r0, #0
8e: d701 bvc.n 94 <stest1>
90: 2301 movs r3, #1
92: 4318 orrs r0, r3
00000094 <stest1>:
94: d301 bcc.n 9a <stest2>
96: 2302 movs r3, #2
98: 4318 orrs r0, r3
0000009a <stest2>:
9a: 4770 bx lr
0000009c <returnvcflags>:
9c: d703 bvc.n a6 <returnvcflags+0xa>
9e: d306 bcc.n ae <returnvcflags+0x12>
a0: 2003 movs r0, #3
a2: 0700 lsls r0, r0, #28
a4: 4770 bx lr
a6: d305 bcc.n b4 <returnvcflags+0x18>
a8: 2002 movs r0, #2
aa: 0700 lsls r0, r0, #28
ac: 4770 bx lr
ae: 2001 movs r0, #1
b0: 0700 lsls r0, r0, #28
b2: 4770 bx lr
b4: 2000 movs r0, #0
b6: 0700 lsls r0, r0, #28
b8: 4770 bx lr
000000ba <testfun1>:
ba: 2200 movs r2, #0
bc: 41d2 rors r2, r2
be: 4148 adcs r0, r1
c0: e7ec b.n 9c <returnvcflags>
000000c2 <testfun2>:
c2: 2201 movs r2, #1
c4: 41d2 rors r2, r2
c6: 4148 adcs r0, r1
c8: e7e8 b.n 9c <returnvcflags>
000000ca <testfun3>:
ca: 2200 movs r2, #0
cc: 41d2 rors r2, r2
ce: 4188 sbcs r0, r1
d0: e7e4 b.n 9c <returnvcflags>
000000d2 <testfun4>:
d2: 2201 movs r2, #1
d4: 41d2 rors r2, r2
d6: 4188 sbcs r0, r1
d8: e7e0 b.n 9c <returnvcflags>
000000da <PUTGETSET>:
da: 6802 ldr r2, [r0, #0]
dc: 430a orrs r2, r1
de: 6002 str r2, [r0, #0]
e0: 4770 bx lr
...
000000e4 <_fini>:
e4: 46c0 nop ; (mov r8, r8)
e6: 4770 bx lr
000000e8 <__exidx_start>:
e8: 46c0 nop ; (mov r8, r8)
ea: 4770 bx lr
000000ec <__exidx_end>:
ec: 46c0 nop ; (mov r8, r8)
ee: 4770 bx lr
000000f0 <__errno>:
f0: 46c0 nop ; (mov r8, r8)
f2: 4770 bx lr
000000f4 <_isatty>:
f4: 2300 movs r3, #0
f6: 0018 movs r0, r3
f8: 4770 bx lr
fa: 46c0 nop ; (mov r8, r8)
000000fc <_fstat>:
fc: 2300 movs r3, #0
fe: 0018 movs r0, r3
100: 4770 bx lr
102: 46c0 nop ; (mov r8, r8)
00000104 <puts>:
104: b500 push {lr}
106: b085 sub sp, #20
108: 9001 str r0, [sp, #4]
10a: 2300 movs r3, #0
10c: 9303 str r3, [sp, #12]
10e: e009 b.n 124 <puts+0x20>
110: 9b03 ldr r3, [sp, #12]
112: 9a01 ldr r2, [sp, #4]
114: 18d3 adds r3, r2, r3
116: 781b ldrb r3, [r3, #0]
118: 0018 movs r0, r3
11a: f7ff ff9e bl 5a <putchar>
11e: 9b03 ldr r3, [sp, #12]
120: 3301 adds r3, #1
122: 9303 str r3, [sp, #12]
124: 9b03 ldr r3, [sp, #12]
126: 9a01 ldr r2, [sp, #4]
128: 18d3 adds r3, r2, r3
12a: 781b ldrb r3, [r3, #0]
12c: 2b00 cmp r3, #0
12e: d1ef bne.n 110 <puts+0xc>
130: 46c0 nop ; (mov r8, r8)
132: b005 add sp, #20
134: bd00 pop {pc}
136: 46c0 nop ; (mov r8, r8)
00000138 <rand>:
138: 2307 movs r3, #7
13a: 0018 movs r0, r3
13c: 4770 bx lr
13e: 46c0 nop ; (mov r8, r8)
00000140 <srand>:
140: 230b movs r3, #11
142: 0018 movs r0, r3
144: 4770 bx lr
146: 46c0 nop ; (mov r8, r8)
00000148 <_write>:
148: b500 push {lr}
14a: b087 sub sp, #28
14c: 9003 str r0, [sp, #12]
14e: 9102 str r1, [sp, #8]
150: 9201 str r2, [sp, #4]
152: 2300 movs r3, #0
154: 9305 str r3, [sp, #20]
156: e00a b.n 16e <_write+0x26>
158: 9b02 ldr r3, [sp, #8]
15a: 781b ldrb r3, [r3, #0]
15c: 0018 movs r0, r3
15e: f7ff ff7c bl 5a <putchar>
162: 9b02 ldr r3, [sp, #8]
164: 3301 adds r3, #1
166: 9302 str r3, [sp, #8]
168: 9b05 ldr r3, [sp, #20]
16a: 3301 adds r3, #1
16c: 9305 str r3, [sp, #20]
16e: 9a05 ldr r2, [sp, #20]
170: 9b01 ldr r3, [sp, #4]
172: 429a cmp r2, r3
174: dbf0 blt.n 158 <_write+0x10>
176: 9b05 ldr r3, [sp, #20]
178: 0018 movs r0, r3
17a: b007 add sp, #28
17c: bd00 pop {pc}
17e: 46c0 nop ; (mov r8, r8)
00000180 <_close>:
180: 46c0 nop ; (mov r8, r8)
182: 4770 bx lr
00000184 <_read>:
184: 46c0 nop ; (mov r8, r8)
186: 4770 bx lr
00000188 <_lseek>:
188: 46c0 nop ; (mov r8, r8)
18a: 4770 bx lr
0000018c <main>:
18c: b500 push {lr}
18e: b085 sub sp, #20
190: 4b06 ldr r3, [pc, #24] ; (1ac <main+0x20>)
192: 9303 str r3, [sp, #12]
194: 4b06 ldr r3, [pc, #24] ; (1b0 <main+0x24>)
196: 9302 str r3, [sp, #8]
198: 9902 ldr r1, [sp, #8]
19a: 9803 ldr r0, [sp, #12]
19c: f000 f80a bl 1b4 <__aeabi_fmul>
1a0: 1c03 adds r3, r0, #0
1a2: 9301 str r3, [sp, #4]
1a4: 2300 movs r3, #0
1a6: 0018 movs r0, r3
1a8: b005 add sp, #20
1aa: bd00 pop {pc}
1ac: 3fcccccd svccc 0x00cccccd
1b0: 400ccccd andmi ip, ip, sp, asr #25
000001b4 <__aeabi_fmul>:
1b4: b5f0 push {r4, r5, r6, r7, lr}
1b6: 4657 mov r7, sl
1b8: 464e mov r6, r9
1ba: 4645 mov r5, r8
1bc: 0043 lsls r3, r0, #1
1be: b4e0 push {r5, r6, r7}
1c0: 0246 lsls r6, r0, #9
1c2: 4688 mov r8, r1
1c4: 0a76 lsrs r6, r6, #9
1c6: 0e1f lsrs r7, r3, #24
1c8: 0fc4 lsrs r4, r0, #31
1ca: 2f00 cmp r7, #0
1cc: d047 beq.n 25e <__aeabi_fmul+0xaa>
1ce: 2fff cmp r7, #255 ; 0xff
1d0: d025 beq.n 21e <__aeabi_fmul+0x6a>
1d2: 2300 movs r3, #0
1d4: 2580 movs r5, #128 ; 0x80
1d6: 469a mov sl, r3
1d8: 4699 mov r9, r3
1da: 00f6 lsls r6, r6, #3
1dc: 04ed lsls r5, r5, #19
1de: 432e orrs r6, r5
1e0: 3f7f subs r7, #127 ; 0x7f
1e2: 4643 mov r3, r8
1e4: 4642 mov r2, r8
1e6: 025d lsls r5, r3, #9
1e8: 0fd2 lsrs r2, r2, #31
1ea: 005b lsls r3, r3, #1
1ec: 0a6d lsrs r5, r5, #9
1ee: 0e1b lsrs r3, r3, #24
1f0: 4690 mov r8, r2
1f2: d040 beq.n 276 <__aeabi_fmul+0xc2>
1f4: 2bff cmp r3, #255 ; 0xff
1f6: d039 beq.n 26c <__aeabi_fmul+0xb8>
1f8: 2280 movs r2, #128 ; 0x80
1fa: 2000 movs r0, #0
1fc: 00ed lsls r5, r5, #3
1fe: 04d2 lsls r2, r2, #19
200: 4315 orrs r5, r2
202: 3b7f subs r3, #127 ; 0x7f
204: 18fb adds r3, r7, r3
206: 4642 mov r2, r8
208: 4657 mov r7, sl
20a: 1c59 adds r1, r3, #1
20c: 4062 eors r2, r4
20e: 468c mov ip, r1
210: 4307 orrs r7, r0
212: 2f0f cmp r7, #15
214: d85c bhi.n 2d0 <__aeabi_fmul+0x11c>
216: 496f ldr r1, [pc, #444] ; (3d4 <__aeabi_fmul+0x220>)
218: 00bf lsls r7, r7, #2
21a: 59c9 ldr r1, [r1, r7]
21c: 468f mov pc, r1
21e: 2e00 cmp r6, #0
220: d145 bne.n 2ae <__aeabi_fmul+0xfa>
222: 2308 movs r3, #8
224: 469a mov sl, r3
226: 3b06 subs r3, #6
228: 4699 mov r9, r3
22a: e7da b.n 1e2 <__aeabi_fmul+0x2e>
22c: 4642 mov r2, r8
22e: 2802 cmp r0, #2
230: d02d beq.n 28e <__aeabi_fmul+0xda>
232: 2803 cmp r0, #3
234: d100 bne.n 238 <__aeabi_fmul+0x84>
236: e0c3 b.n 3c0 <__aeabi_fmul+0x20c>
238: 2801 cmp r0, #1
23a: d000 beq.n 23e <__aeabi_fmul+0x8a>
23c: e0a2 b.n 384 <__aeabi_fmul+0x1d0>
23e: 2500 movs r5, #0
240: 2600 movs r6, #0
242: 4002 ands r2, r0
244: b2d4 uxtb r4, r2
246: 0276 lsls r6, r6, #9
248: 05ed lsls r5, r5, #23
24a: 0a76 lsrs r6, r6, #9
24c: 432e orrs r6, r5
24e: 07e4 lsls r4, r4, #31
250: 4326 orrs r6, r4
252: 0030 movs r0, r6
254: bc1c pop {r2, r3, r4}
256: 4690 mov r8, r2
258: 4699 mov r9, r3
25a: 46a2 mov sl, r4
25c: bdf0 pop {r4, r5, r6, r7, pc}
25e: 2e00 cmp r6, #0
260: d11a bne.n 298 <__aeabi_fmul+0xe4>
262: 2304 movs r3, #4
264: 469a mov sl, r3
266: 3b03 subs r3, #3
268: 4699 mov r9, r3
26a: e7ba b.n 1e2 <__aeabi_fmul+0x2e>
26c: 002a movs r2, r5
26e: 1e51 subs r1, r2, #1
270: 418a sbcs r2, r1
272: 1c90 adds r0, r2, #2
274: e7c6 b.n 204 <__aeabi_fmul+0x50>
276: 2001 movs r0, #1
278: 2d00 cmp r5, #0
27a: d0c3 beq.n 204 <__aeabi_fmul+0x50>
27c: 0028 movs r0, r5
27e: f000 f8ad bl 3dc <__clzsi2>
282: 1f43 subs r3, r0, #5
284: 3076 adds r0, #118 ; 0x76
286: 409d lsls r5, r3
288: 4243 negs r3, r0
28a: 2000 movs r0, #0
28c: e7ba b.n 204 <__aeabi_fmul+0x50>
28e: 2401 movs r4, #1
290: 25ff movs r5, #255 ; 0xff
292: 4014 ands r4, r2
294: 2600 movs r6, #0
296: e7d6 b.n 246 <__aeabi_fmul+0x92>
298: 0030 movs r0, r6
29a: f000 f89f bl 3dc <__clzsi2>
29e: 1f43 subs r3, r0, #5
2a0: 409e lsls r6, r3
2a2: 2300 movs r3, #0
2a4: 3076 adds r0, #118 ; 0x76
2a6: 4247 negs r7, r0
2a8: 469a mov sl, r3
2aa: 4699 mov r9, r3
2ac: e799 b.n 1e2 <__aeabi_fmul+0x2e>
2ae: 230c movs r3, #12
2b0: 469a mov sl, r3
2b2: 3b09 subs r3, #9
2b4: 4699 mov r9, r3
2b6: e794 b.n 1e2 <__aeabi_fmul+0x2e>
2b8: 2680 movs r6, #128 ; 0x80
2ba: 2400 movs r4, #0
2bc: 03f6 lsls r6, r6, #15
2be: 25ff movs r5, #255 ; 0xff
2c0: e7c1 b.n 246 <__aeabi_fmul+0x92>
2c2: 0035 movs r5, r6
2c4: 4648 mov r0, r9
2c6: e7b2 b.n 22e <__aeabi_fmul+0x7a>
2c8: 0035 movs r5, r6
2ca: 0022 movs r2, r4
2cc: 4648 mov r0, r9
2ce: e7ae b.n 22e <__aeabi_fmul+0x7a>
2d0: 0429 lsls r1, r5, #16
2d2: 0c09 lsrs r1, r1, #16
2d4: 0008 movs r0, r1
2d6: 0c37 lsrs r7, r6, #16
2d8: 0436 lsls r6, r6, #16
2da: 0c36 lsrs r6, r6, #16
2dc: 0c2c lsrs r4, r5, #16
2de: 4379 muls r1, r7
2e0: 4370 muls r0, r6
2e2: 4367 muls r7, r4
2e4: 4374 muls r4, r6
2e6: 0c06 lsrs r6, r0, #16
2e8: 1864 adds r4, r4, r1
2ea: 1936 adds r6, r6, r4
2ec: 42b1 cmp r1, r6
2ee: d903 bls.n 2f8 <__aeabi_fmul+0x144>
2f0: 2180 movs r1, #128 ; 0x80
2f2: 0249 lsls r1, r1, #9
2f4: 4688 mov r8, r1
2f6: 4447 add r7, r8
2f8: 0400 lsls r0, r0, #16
2fa: 0c00 lsrs r0, r0, #16
2fc: 0431 lsls r1, r6, #16
2fe: 1809 adds r1, r1, r0
300: 018d lsls r5, r1, #6
302: 1e68 subs r0, r5, #1
304: 4185 sbcs r5, r0
306: 0e89 lsrs r1, r1, #26
308: 4329 orrs r1, r5
30a: 0c35 lsrs r5, r6, #16
30c: 19ed adds r5, r5, r7
30e: 01ad lsls r5, r5, #6
310: 430d orrs r5, r1
312: 0129 lsls r1, r5, #4
314: d504 bpl.n 320 <__aeabi_fmul+0x16c>
316: 2301 movs r3, #1
318: 0869 lsrs r1, r5, #1
31a: 401d ands r5, r3
31c: 4663 mov r3, ip
31e: 430d orrs r5, r1
320: 0019 movs r1, r3
322: 317f adds r1, #127 ; 0x7f
324: 2900 cmp r1, #0
326: dd25 ble.n 374 <__aeabi_fmul+0x1c0>
328: 0768 lsls r0, r5, #29
32a: d004 beq.n 336 <__aeabi_fmul+0x182>
32c: 200f movs r0, #15
32e: 4028 ands r0, r5
330: 2804 cmp r0, #4
332: d000 beq.n 336 <__aeabi_fmul+0x182>
334: 3504 adds r5, #4
336: 0128 lsls r0, r5, #4
338: d503 bpl.n 342 <__aeabi_fmul+0x18e>
33a: 4927 ldr r1, [pc, #156] ; (3d8 <__aeabi_fmul+0x224>)
33c: 3380 adds r3, #128 ; 0x80
33e: 400d ands r5, r1
340: 0019 movs r1, r3
342: 29fe cmp r1, #254 ; 0xfe
344: dca3 bgt.n 28e <__aeabi_fmul+0xda>
346: 2401 movs r4, #1
348: 01ad lsls r5, r5, #6
34a: 0a6e lsrs r6, r5, #9
34c: 4014 ands r4, r2
34e: b2cd uxtb r5, r1
350: e779 b.n 246 <__aeabi_fmul+0x92>
352: 2080 movs r0, #128 ; 0x80
354: 03c0 lsls r0, r0, #15
356: 4206 tst r6, r0
358: d007 beq.n 36a <__aeabi_fmul+0x1b6>
35a: 4205 tst r5, r0
35c: d105 bne.n 36a <__aeabi_fmul+0x1b6>
35e: 4328 orrs r0, r5
360: 0246 lsls r6, r0, #9
362: 0a76 lsrs r6, r6, #9
364: 4644 mov r4, r8
366: 25ff movs r5, #255 ; 0xff
368: e76d b.n 246 <__aeabi_fmul+0x92>
36a: 4306 orrs r6, r0
36c: 0276 lsls r6, r6, #9
36e: 0a76 lsrs r6, r6, #9
370: 25ff movs r5, #255 ; 0xff
372: e768 b.n 246 <__aeabi_fmul+0x92>
374: 2401 movs r4, #1
376: 1a61 subs r1, r4, r1
378: 291b cmp r1, #27
37a: dd05 ble.n 388 <__aeabi_fmul+0x1d4>
37c: 4014 ands r4, r2
37e: 2500 movs r5, #0
380: 2600 movs r6, #0
382: e760 b.n 246 <__aeabi_fmul+0x92>
384: 4663 mov r3, ip
386: e7cb b.n 320 <__aeabi_fmul+0x16c>
388: 002e movs r6, r5
38a: 2320 movs r3, #32
38c: 40ce lsrs r6, r1
38e: 1a59 subs r1, r3, r1
390: 408d lsls r5, r1
392: 1e6b subs r3, r5, #1
394: 419d sbcs r5, r3
396: 432e orrs r6, r5
398: 0773 lsls r3, r6, #29
39a: d004 beq.n 3a6 <__aeabi_fmul+0x1f2>
39c: 230f movs r3, #15
39e: 4033 ands r3, r6
3a0: 2b04 cmp r3, #4
3a2: d000 beq.n 3a6 <__aeabi_fmul+0x1f2>
3a4: 3604 adds r6, #4
3a6: 0173 lsls r3, r6, #5
3a8: d504 bpl.n 3b4 <__aeabi_fmul+0x200>
3aa: 2401 movs r4, #1
3ac: 2501 movs r5, #1
3ae: 4014 ands r4, r2
3b0: 2600 movs r6, #0
3b2: e748 b.n 246 <__aeabi_fmul+0x92>
3b4: 2401 movs r4, #1
3b6: 01b6 lsls r6, r6, #6
3b8: 0a76 lsrs r6, r6, #9
3ba: 4014 ands r4, r2
3bc: 2500 movs r5, #0
3be: e742 b.n 246 <__aeabi_fmul+0x92>
3c0: 2680 movs r6, #128 ; 0x80
3c2: 2401 movs r4, #1
3c4: 03f6 lsls r6, r6, #15
3c6: 432e orrs r6, r5
3c8: 0276 lsls r6, r6, #9
3ca: 0a76 lsrs r6, r6, #9
3cc: 4014 ands r4, r2
3ce: 25ff movs r5, #255 ; 0xff
3d0: e739 b.n 246 <__aeabi_fmul+0x92>
3d2: 46c0 nop ; (mov r8, r8)
3d4: 00000418 andeq r0, r0, r8, lsl r4
3d8: f7ffffff ; <UNDEFINED> instruction: 0xf7ffffff
000003dc <__clzsi2>:
3dc: 211c movs r1, #28
3de: 2301 movs r3, #1
3e0: 041b lsls r3, r3, #16
3e2: 4298 cmp r0, r3
3e4: d301 bcc.n 3ea <__clzsi2+0xe>
3e6: 0c00 lsrs r0, r0, #16
3e8: 3910 subs r1, #16
3ea: 0a1b lsrs r3, r3, #8
3ec: 4298 cmp r0, r3
3ee: d301 bcc.n 3f4 <__clzsi2+0x18>
3f0: 0a00 lsrs r0, r0, #8
3f2: 3908 subs r1, #8
3f4: 091b lsrs r3, r3, #4
3f6: 4298 cmp r0, r3
3f8: d301 bcc.n 3fe <__clzsi2+0x22>
3fa: 0900 lsrs r0, r0, #4
3fc: 3904 subs r1, #4
3fe: a202 add r2, pc, #8 ; (adr r2, 408 <__clzsi2+0x2c>)
400: 5c10 ldrb r0, [r2, r0]
402: 1840 adds r0, r0, r1
404: 4770 bx lr
406: 46c0 nop ; (mov r8, r8)
408: 02020304 andeq r0, r2, #4, 6 ; 0x10000000
40c: 01010101 tsteq r1, r1, lsl #2
...
Disassembly of section .rodata:
00000418 <.rodata>:
418: 000002d0 ldrdeq r0, [r0], -r0 ; <UNPREDICTABLE>
41c: 0000022e andeq r0, r0, lr, lsr #4
420: 0000022e andeq r0, r0, lr, lsr #4
424: 0000022c andeq r0, r0, ip, lsr #4
428: 000002c2 andeq r0, r0, r2, asr #5
42c: 000002c2 andeq r0, r0, r2, asr #5
430: 000002b8 ; <UNDEFINED> instruction: 0x000002b8
434: 0000022c andeq r0, r0, ip, lsr #4
438: 000002c2 andeq r0, r0, r2, asr #5
43c: 000002b8 ; <UNDEFINED> instruction: 0x000002b8
440: 000002c2 andeq r0, r0, r2, asr #5
444: 0000022c andeq r0, r0, ip, lsr #4
448: 000002c8 andeq r0, r0, r8, asr #5
44c: 000002c8 andeq r0, r0, r8, asr #5
450: 000002c8 andeq r0, r0, r8, asr #5
454: 00000352 andeq r0, r0, r2, asr r3
Disassembly of section .data:
00000458 <heapCurrent>:
458: 40000000 andmi r0, r0, r0
Disassembly of section .ARM.attributes:
00000000 <.ARM.attributes>:
0: 00002d41 andeq r2, r0, r1, asr #26
4: 61656100 cmnvs r5, r0, lsl #2
8: 01006962 tsteq r0, r2, ror #18
c: 00000023 andeq r0, r0, r3, lsr #32
10: 726f4305 rsbvc r4, pc, #335544320 ; 0x14000000
14: 2d786574 cfldr64cs mvdx6, [r8, #-464]! ; 0xfffffe30
18: 002b304d eoreq r3, fp, sp, asr #32
1c: 4d070c06 stcmi 12, cr0, [r7, #-24] ; 0xffffffe8
20: 04120109 ldreq r0, [r2], #-265 ; 0xfffffef7
24: 01150114 tsteq r5, r4, lsl r1
28: 01180317 tsteq r8, r7, lsl r3
2c: Address 0x0000002c is out of bounds.
Disassembly of section .comment:
00000000 <.comment>:
0: 3a434347 bcc 10d0d24 <heapCurrent+0x10d08cc>
4: 4e472820 cdpmi 8, 4, cr2, cr7, cr0, {1}
8: 6f542055 svcvs 0x00542055
c: 20736c6f rsbscs r6, r3, pc, ror #24
10: 20726f66 rsbscs r6, r2, r6, ror #30
14: 204d5241 subcs r5, sp, r1, asr #4
18: 65626d45 strbvs r6, [r2, #-3397]! ; 0xfffff2bb
1c: 64656464 strbtvs r6, [r5], #-1124 ; 0xfffffb9c
20: 6f725020 svcvs 0x00725020
24: 73736563 cmnvc r3, #415236096 ; 0x18c00000
28: 2973726f ldmdbcs r3!, {r0, r1, r2, r3, r5, r6, r9, ip, sp, lr}^
2c: 322e3520 eorcc r3, lr, #32, 10 ; 0x8000000
30: 3220312e eorcc r3, r0, #-2147483637 ; 0x8000000b
34: 31353130 teqcc r5, r0, lsr r1
38: 20323032 eorscs r3, r2, r2, lsr r0
3c: 6c657228 sfmvs f7, 2, [r5], #-160 ; 0xffffff60
40: 65736165 ldrbvs r6, [r3, #-357]! ; 0xfffffe9b
44: 415b2029 cmpmi fp, r9, lsr #32
48: 652f4d52 strvs r4, [pc, #-3410]! ; fffff2fe <_stacktop+0xbf7ff2fe>
4c: 6465626d strbtvs r6, [r5], #-621 ; 0xfffffd93
50: 2d646564 cfstr64cs mvdx6, [r4, #-400]! ; 0xfffffe70
54: 72622d35 rsbvc r2, r2, #3392 ; 0xd40
58: 68636e61 stmdavs r3!, {r0, r5, r6, r9, sl, fp, sp, lr}^
5c: 76657220 strbtvc r7, [r5], -r0, lsr #4
60: 6f697369 svcvs 0x00697369
64: 3332206e teqcc r2, #110 ; 0x6e
68: 38343831 ldmdacc r4!, {r0, r4, r5, fp, ip, sp}
6c: Address 0x0000006c is out of bounds.
Disassembly of section .debug_frame:
00000000 <.debug_frame>:
0: 0000000c andeq r0, r0, ip
4: ffffffff ; <UNDEFINED> instruction: 0xffffffff
8: 7c020001 stcvc 0, cr0, [r2], {1}
c: 000d0c0e andeq r0, sp, lr, lsl #24
10: 00000024 andeq r0, r0, r4, lsr #32
14: 00000000 andeq r0, r0, r0
18: 000001b4 ; <UNDEFINED> instruction: 0x000001b4
1c: 00000228 andeq r0, r0, r8, lsr #4
20: 84140e41 ldrhi r0, [r4], #-3649 ; 0xfffff1bf
24: 86048505 strhi r8, [r4], -r5, lsl #10
28: 8e028703 cdphi 7, 0, cr8, cr2, cr3, {0}
2c: 200e4501 andcs r4, lr, r1, lsl #10
30: 07890888 streq r0, [r9, r8, lsl #17]
34: 0000068a andeq r0, r0, sl, lsl #13
Disassembly of non-working code:
test.elf: file format elf32-littlearm
Disassembly of section .text:
... (all the code from 0x0 to 0x40C are identical. I just created explicit labels within the last two functions to help me with what im trying to do!)
000003fe <twentynine>:
3fe: a202 add r2, pc, #8 ; (adr r2, 408 <twentynine+0xa>)
400: 5c10 ldrb r0, [r2, r0]
402: 1840 adds r0, r0, r1
404: 4770 bx lr
406: 46c0 nop ; (mov r8, r8)
408: 02020304 andeq r0, r2, #4, 6 ; 0x10000000
40c: 01010101 tsteq r1, r1, lsl #2
Disassembly of section .data:
00000410 <heapCurrent>:
410: 40000000 andmi r0, r0, r0
Disassembly of section .ARM.attributes:
00000000 <.ARM.attributes>:
0: 00002d41 andeq r2, r0, r1, asr #26
4: 61656100 cmnvs r5, r0, lsl #2
8: 01006962 tsteq r0, r2, ror #18
c: 00000023 andeq r0, r0, r3, lsr #32
10: 726f4305 rsbvc r4, pc, #335544320 ; 0x14000000
14: 2d786574 cfldr64cs mvdx6, [r8, #-464]! ; 0xfffffe30
18: 002b304d eoreq r3, fp, sp, asr #32
1c: 4d070c06 stcmi 12, cr0, [r7, #-24] ; 0xffffffe8
20: 04120109 ldreq r0, [r2], #-265 ; 0xfffffef7
24: 01150114 tsteq r5, r4, lsl r1
28: 01180317 tsteq r8, r7, lsl r3
2c: Address 0x0000002c is out of bounds.
Disassembly of section .comment:
00000000 <.comment>:
0: 3a434347 bcc 10d0d24 <heapCurrent+0x10d0914>
4: 4e472820 cdpmi 8, 4, cr2, cr7, cr0, {1}
8: 6f542055 svcvs 0x00542055
c: 20736c6f rsbscs r6, r3, pc, ror #24
10: 20726f66 rsbscs r6, r2, r6, ror #30
14: 204d5241 subcs r5, sp, r1, asr #4
18: 65626d45 strbvs r6, [r2, #-3397]! ; 0xfffff2bb
1c: 64656464 strbtvs r6, [r5], #-1124 ; 0xfffffb9c
20: 6f725020 svcvs 0x00725020
24: 73736563 cmnvc r3, #415236096 ; 0x18c00000
28: 2973726f ldmdbcs r3!, {r0, r1, r2, r3, r5, r6, r9, ip, sp, lr}^
2c: 322e3520 eorcc r3, lr, #32, 10 ; 0x8000000
30: 3220312e eorcc r3, r0, #-2147483637 ; 0x8000000b
34: 31353130 teqcc r5, r0, lsr r1
38: 20323032 eorscs r3, r2, r2, lsr r0
3c: 6c657228 sfmvs f7, 2, [r5], #-160 ; 0xffffff60
40: 65736165 ldrbvs r6, [r3, #-357]! ; 0xfffffe9b
44: 415b2029 cmpmi fp, r9, lsr #32
48: 652f4d52 strvs r4, [pc, #-3410]! ; fffff2fe <_stacktop+0xbf7ff2fe>
4c: 6465626d strbtvs r6, [r5], #-621 ; 0xfffffd93
50: 2d646564 cfstr64cs mvdx6, [r4, #-400]! ; 0xfffffe70
54: 72622d35 rsbvc r2, r2, #3392 ; 0xd40
58: 68636e61 stmdavs r3!, {r0, r5, r6, r9, sl, fp, sp, lr}^
5c: 76657220 strbtvc r7, [r5], -r0, lsr #4
60: 6f697369 svcvs 0x00697369
64: 3332206e teqcc r2, #110 ; 0x6e
68: 38343831 ldmdacc r4!, {r0, r4, r5, fp, ip, sp}
6c: Address 0x0000006c is out of bounds.

Replacing memcpy and .divsi3_skip_div0_test with smaller code on ARM MCU

My entry for https://hackaday.com/2016/11/21/step-up-to-the-1-kb-challenge/ includes a couple of huge functions which are not generated from any C code which I have written.
000004e4 <.divsi3_skip_div0_test>:
4e4: b410 push {r4}
4e6: 1c04 adds r4, r0, #0
4e8: 404c eors r4, r1
4ea: 46a4 mov ip, r4
4ec: 2301 movs r3, #1
4ee: 2200 movs r2, #0
4f0: 2900 cmp r1, #0
4f2: d500 bpl.n 4f6 <.divsi3_skip_div0_test+0x12>
4f4: 4249 negs r1, r1
4f6: 2800 cmp r0, #0
4f8: d500 bpl.n 4fc <.divsi3_skip_div0_test+0x18>
4fa: 4240 negs r0, r0
4fc: 4288 cmp r0, r1
4fe: d32c bcc.n 55a <.divsi3_skip_div0_test+0x76>
500: 2401 movs r4, #1
502: 0724 lsls r4, r4, #28
504: 42a1 cmp r1, r4
506: d204 bcs.n 512 <.divsi3_skip_div0_test+0x2e>
508: 4281 cmp r1, r0
50a: d202 bcs.n 512 <.divsi3_skip_div0_test+0x2e>
50c: 0109 lsls r1, r1, #4
50e: 011b lsls r3, r3, #4
510: e7f8 b.n 504 <.divsi3_skip_div0_test+0x20>
512: 00e4 lsls r4, r4, #3
514: 42a1 cmp r1, r4
516: d204 bcs.n 522 <.divsi3_skip_div0_test+0x3e>
518: 4281 cmp r1, r0
51a: d202 bcs.n 522 <.divsi3_skip_div0_test+0x3e>
51c: 0049 lsls r1, r1, #1
51e: 005b lsls r3, r3, #1
520: e7f8 b.n 514 <.divsi3_skip_div0_test+0x30>
522: 4288 cmp r0, r1
524: d301 bcc.n 52a <.divsi3_skip_div0_test+0x46>
526: 1a40 subs r0, r0, r1
528: 431a orrs r2, r3
52a: 084c lsrs r4, r1, #1
52c: 42a0 cmp r0, r4
52e: d302 bcc.n 536 <.divsi3_skip_div0_test+0x52>
530: 1b00 subs r0, r0, r4
532: 085c lsrs r4, r3, #1
534: 4322 orrs r2, r4
536: 088c lsrs r4, r1, #2
538: 42a0 cmp r0, r4
53a: d302 bcc.n 542 <.divsi3_skip_div0_test+0x5e>
53c: 1b00 subs r0, r0, r4
53e: 089c lsrs r4, r3, #2
540: 4322 orrs r2, r4
542: 08cc lsrs r4, r1, #3
544: 42a0 cmp r0, r4
546: d302 bcc.n 54e <.divsi3_skip_div0_test+0x6a>
548: 1b00 subs r0, r0, r4
54a: 08dc lsrs r4, r3, #3
54c: 4322 orrs r2, r4
54e: 2800 cmp r0, #0
550: d003 beq.n 55a <.divsi3_skip_div0_test+0x76>
552: 091b lsrs r3, r3, #4
554: d001 beq.n 55a <.divsi3_skip_div0_test+0x76>
556: 0909 lsrs r1, r1, #4
558: e7e3 b.n 522 <.divsi3_skip_div0_test+0x3e>
55a: 1c10 adds r0, r2, #0
55c: 4664 mov r4, ip
55e: 2c00 cmp r4, #0
560: d500 bpl.n 564 <.divsi3_skip_div0_test+0x80>
562: 4240 negs r0, r0
564: bc10 pop {r4}
566: 4770 bx lr
568: 2800 cmp r0, #0
56a: d006 beq.n 57a <.divsi3_skip_div0_test+0x96>
56c: db03 blt.n 576 <.divsi3_skip_div0_test+0x92>
56e: 2000 movs r0, #0
570: 43c0 mvns r0, r0
572: 0840 lsrs r0, r0, #1
574: e001 b.n 57a <.divsi3_skip_div0_test+0x96>
576: 2080 movs r0, #128 ; 0x80
578: 0600 lsls r0, r0, #24
57a: b407 push {r0, r1, r2}
57c: 4802 ldr r0, [pc, #8] ; (588 <.divsi3_skip_div0_test+0xa4>)
57e: a102 add r1, pc, #8 ; (adr r1, 588 <.divsi3_skip_div0_test+0xa4>)
580: 1840 adds r0, r0, r1
582: 9002 str r0, [sp, #8]
584: bd03 pop {r0, r1, pc}
586: 46c0 nop ; (mov r8, r8)
588: 00000019 .word 0x00000019
and:
000005a4 <memcpy>:
5a4: b5f0 push {r4, r5, r6, r7, lr}
5a6: 2a0f cmp r2, #15
5a8: d935 bls.n 616 <memcpy+0x72>
5aa: 1c03 adds r3, r0, #0
5ac: 430b orrs r3, r1
5ae: 079c lsls r4, r3, #30
5b0: d135 bne.n 61e <memcpy+0x7a>
5b2: 1c16 adds r6, r2, #0
5b4: 3e10 subs r6, #16
5b6: 0936 lsrs r6, r6, #4
5b8: 0135 lsls r5, r6, #4
5ba: 1945 adds r5, r0, r5
5bc: 3510 adds r5, #16
5be: 1c0c adds r4, r1, #0
5c0: 1c03 adds r3, r0, #0
5c2: 6827 ldr r7, [r4, #0]
5c4: 601f str r7, [r3, #0]
5c6: 6867 ldr r7, [r4, #4]
5c8: 605f str r7, [r3, #4]
5ca: 68a7 ldr r7, [r4, #8]
5cc: 609f str r7, [r3, #8]
5ce: 68e7 ldr r7, [r4, #12]
5d0: 3410 adds r4, #16
5d2: 60df str r7, [r3, #12]
5d4: 3310 adds r3, #16
5d6: 42ab cmp r3, r5
5d8: d1f3 bne.n 5c2 <memcpy+0x1e>
5da: 1c73 adds r3, r6, #1
5dc: 011b lsls r3, r3, #4
5de: 18c5 adds r5, r0, r3
5e0: 18c9 adds r1, r1, r3
5e2: 230f movs r3, #15
5e4: 4013 ands r3, r2
5e6: 2b03 cmp r3, #3
5e8: d91b bls.n 622 <memcpy+0x7e>
5ea: 1f1c subs r4, r3, #4
5ec: 08a4 lsrs r4, r4, #2
5ee: 3401 adds r4, #1
5f0: 00a4 lsls r4, r4, #2
5f2: 2300 movs r3, #0
5f4: 58ce ldr r6, [r1, r3]
5f6: 50ee str r6, [r5, r3]
5f8: 3304 adds r3, #4
5fa: 42a3 cmp r3, r4
5fc: d1fa bne.n 5f4 <memcpy+0x50>
5fe: 18ed adds r5, r5, r3
600: 18c9 adds r1, r1, r3
602: 2303 movs r3, #3
604: 401a ands r2, r3
606: d005 beq.n 614 <memcpy+0x70>
608: 2300 movs r3, #0
60a: 5ccc ldrb r4, [r1, r3]
60c: 54ec strb r4, [r5, r3]
60e: 3301 adds r3, #1
610: 4293 cmp r3, r2
612: d1fa bne.n 60a <memcpy+0x66>
614: bdf0 pop {r4, r5, r6, r7, pc}
616: 1c05 adds r5, r0, #0
618: 2a00 cmp r2, #0
61a: d1f5 bne.n 608 <memcpy+0x64>
61c: e7fa b.n 614 <memcpy+0x70>
61e: 1c05 adds r5, r0, #0
620: e7f2 b.n 608 <memcpy+0x64>
622: 1c1a adds r2, r3, #0
624: e7f8 b.n 618 <memcpy+0x74>
626: 46c0 nop ; (mov r8, r8)
I am guessing that I could code far smaller, but less time-efficient, replacements myself.
Is this likely?
Where will I find the source which I need to edit - I'm guessing that I should look for the source of a libc under gcc-arm-none-eabi/lib/gcc/arm-none-eabi/4.8.3/. I think that I've found the compiled symbols, but I can't find the source.
~/gcc-arm-none-eabi$ grep -R divsi3_skip_div0_test *
Binary file lib/gcc/arm-none-eabi/4.8.3/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/thumb/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/armv6-m/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/fpu/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/armv7-ar/thumb/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/armv7-ar/thumb/softfp/libgcc.a matches
Binary file lib/gcc/arm-none-eabi/4.8.3/armv7-ar/thumb/fpu/libgcc.a matches
Alternatively, is there a way to tell gcc not to use memcpy when copying structures? (They are 10 bytes, so three thumb instructions should do the job.) I've tried adding -mno-memcpy and -Wa,mno-memcpy but neither are recognised.
Update:
I've solved the memcpy part of this question - adding a partial, but sufficient, memcpy function stops the other from being added.
size_t memcpy(uint8_t *restrict dst, uint8_t *restrict const src, size_t size) {
int i;
for (i = 0; i < size; i++) {
dst[i] = src[i];
}
return i;
}
It's much smaller, but less efficient and won't handle dst < src + size overlap.
000003ec <memcpy>:
3ec: b510 push {r4, lr}
3ee: 2300 movs r3, #0
3f0: 4293 cmp r3, r2
3f2: d003 beq.n 3fc <memcpy+0x10>
3f4: 5ccc ldrb r4, [r1, r3]
3f6: 54c4 strb r4, [r0, r3]
3f8: 3301 adds r3, #1
3fa: e7f9 b.n 3f0 <memcpy+0x4>
3fc: 1c18 adds r0, r3, #0
3fe: bd10 pop {r4, pc}
To clarify, I'm now only asking what I might do to replace the .divsi3_skip_div0_test code with a less efficient, but smaller, code.
It is not clear to me where the source of this code is, or how to edit its source. It looks to be more complicated to replace than memcpy as it does not look like a C function, as its name begins with a ..

What could be the possible reason that an ARM register gets polluted?

Recently I ran into this strange issue(Thumb2 instr-set). Say I have following code:
000a8948 <some_func>:
a8948: e92d 4ff0 stmdb sp!, {r4, r5, r6, r7, r8, r9, sl, fp, lr}
a894c: 460f mov r7, r1
a894e: f8df 8260 ldr.w r8, [pc, #608]
a8952: 4606 mov r6, r0 --> r0 shows b89a8408 in backtrace
a8954: 4b97 ldr r3, [pc, #604]
a8956: f100 0410 add.w r4, r0, #16 --> only place r4 gets assigned, but it can't be 0.
a895a: 44f8 add r8, pc
a895c: f04f 0900 mov.w r9, #0
a8960: ed2d 8b04 vpush {d8-d9}
a8964: f858 1003 ldr.w r1, [r8, r3]
a8968: b095 sub sp, #84 ; 0x54
a896a: f8df b24c ldr.w fp, [pc, #588]
a896e: ad0c add r5, sp, #48
a8970: f8df c248 ldr.w ip, [pc, #584]
a8974: f8df a248 ldr.w sl, [pc, #584]
a8978: ed9f 8b8b vldr d8, [pc, #556]
a897c: 680a ldr r2, [r1, #0]
a897e: 44fb add fp, pc
a8980: 44fc add ip, pc
a8982: 44fa add sl, pc
a8984: 9107 str r1, [sp, #28]
a8986: 9213 str r2, [sp, #76]
a8988: 6823 ldr r3, [r4, #0] // ERR: SIGSEGV happens, code 1 (SEGV_MAPPER), fault addr 0x0
So when this issue happens, r4 turned to value 0, what I cannot figure out is that how r4 gets polluted, given it only be assigned to r0+16 at line a8956? Could it be that Neon instructions write r4 somehow?

simple ADD/ADC ARM assemlby fails

I have the following C and ASM version of the (supposedly) same code. What it does is load 2 128bit ints represented by 2 64bit ints each to registers (first 4*lower 32bit, then 4*higher 32bit) and ADD/ADC to them. It is simple enough code and the ARM/ST manuals actually give the same example with 96bit (3 ADD/ADCs).
For simple calls both versions work (repeatedly adding (1 << x++) or 1..x). But for the longer testsuite the ARM assembly fails (board hangs). ATM I have no ability to trap/debug that and cannot use any printf() or the likes to find the test failing, which is irrelevant anyways, because there must be some basic fault in the ASM version, as the C version works as expected.
I don't get it, it's simple enough and very close to the C assembly output (sans branching). I tried the "memory" constraint (shouldn't be needed), I tried saving the carry between lower and upper 64bit in a register and adding that later, using ADD(C).W, alignment, using two LDR/STR instead of LDRD/STRD, etc.. I assume the board faults because some addition goes wrong and results in a divide by 0 or something like that.
The GCC ASM is below and uses similar basic technique, so I don't see the problem.
I'm really just looking for the fastest way to do the add, not to fix that code specifically. It's a shame you have to use constant register names because there is no constraint for specifying rX and rX+1. Also it's impossible to use as many registers as GCC as it will run out of them during compilation.
typedef struct I128 {
int64_t high;
uint64_t low;
} I128;
I128 I128add(I128 a, const I128 b) {
#if defined(USEASM) && defined(ARMx)
__asm(
"LDRD %%r2, %%r3, %[alo]\n"
"LDRD %%r4, %%r5, %[blo]\n"
"ADDS %%r2, %%r2, %%r4\n"
"ADCS %%r3, %%r3, %%r5\n"
"STRD %%r2, %%r3, %[alo]\n"
"LDRD %%r2, %%r3, %[ahi]\n"
"LDRD %%r4, %%r5, %[bhi]\n"
"ADCS %%r2, %%r2, %%r4\n"
"ADC %%r3, %%r3, %%r5\n"
"STRD %%r2, %%r3, %[ahi]\n"
: [alo] "+m" (a.low), [ahi] "+m" (a.high)
: [blo] "m" (b.low), [bhi] "m" (b.high)
: "r2", "r3", "r4", "r5", "cc"
);
return a;
#else
// faster to use temp than saving low and adding to a directly
I128 r = {a.high + b.high, a.low + b.low};
// check for overflow of low 64 bits, add carry to high
// avoid conditionals
//r.high += r.low < a.low || r.low < b.low;
// actually gcc produces faster code with conditionals
if(r.low < a.low || r.low < b.low) ++r.high;
return r;
}
GCC C version using " armv7m-none-eabi-gcc-4.7.2 -O3 -ggdb -fomit-frame-pointer -falign-functions=16 -std=gnu99 -march=armv7e-m":
b082 sub sp, #8
e92d 0ff0 stmdb sp!, {r4, r5, r6, r7, r8, r9, sl, fp}
a908 add r1, sp, #32
e881 000c stmia.w r1, {r2, r3}
e9dd 890e ldrd r8, r9, [sp, #56] ; 0x38
e9dd 670a ldrd r6, r7, [sp, #40] ; 0x28
e9dd 2308 ldrd r2, r3, [sp, #32]
e9dd 450c ldrd r4, r5, [sp, #48] ; 0x30
eb16 0a08 adds.w sl, r6, r8
eb47 0b09 adc.w fp, r7, r9
1912 adds r2, r2, r4
eb43 0305 adc.w r3, r3, r5
45bb cmp fp, r7
bf08 it eq
45b2 cmpeq sl, r6
d303 bcc.n 8012c9a <I128add+0x3a>
45cb cmp fp, r9
bf08 it eq
45c2 cmpeq sl, r8
d204 bcs.n 8012ca4 <I128add+0x44>
2401 movs r4, #1
2500 movs r5, #0
1912 adds r2, r2, r4
eb43 0305 adc.w r3, r3, r5
e9c0 2300 strd r2, r3, [r0]
e9c0 ab02 strd sl, fp, [r0, #8]
e8bd 0ff0 ldmia.w sp!, {r4, r5, r6, r7, r8, r9, sl, fp}
b002 add sp, #8
4770 bx lr
My ASM version that fails:
b082 sub sp, #8
b430 push {r4, r5}
a902 add r1, sp, #8
e881 000c stmia.w r1, {r2, r3}
e9dd 2304 ldrd r2, r3, [sp, #16]
e9dd 4508 ldrd r4, r5, [sp, #32]
1912 adds r2, r2, r4
416b adcs r3, r5
e9cd 2304 strd r2, r3, [sp, #16]
e9dd 2302 ldrd r2, r3, [sp, #8]
e9dd 4506 ldrd r4, r5, [sp, #24]
4162 adcs r2, r4
eb43 0305 adc.w r3, r3, r5
e9cd 2302 strd r2, r3, [sp, #8]
4604 mov r4, r0
c90f ldmia r1, {r0, r1, r2, r3}
e884 000f stmia.w r4, {r0, r1, r2, r3}
4620 mov r0, r4
bc30 pop {r4, r5}
b002 add sp, #8
4770 bx lr
I am not getting a hang from your code, but it isnt working either, not sure why. But it was very easy to patch the compiler generated code to handle the carry:
I128 I128add(I128 a, const I128 b) {
I128 r = {a.high + b.high, a.low + b.low};
return r;
}
becomes
000001e4 <I128add>:
1e4: b082 sub sp, #8
1e6: b4f0 push {r4, r5, r6, r7}
1e8: e9dd 4506 ldrd r4, r5, [sp, #24]
1ec: a904 add r1, sp, #16
1ee: e881 000c stmia.w r1, {r2, r3}
1f2: e9dd 230a ldrd r2, r3, [sp, #40] ; 0x28
1f6: 1912 adds r2, r2, r4
1f8: eb43 0305 adc.w r3, r3, r5
1fc: e9dd 6704 ldrd r6, r7, [sp, #16]
200: e9dd 4508 ldrd r4, r5, [sp, #32]
204: 1936 adds r6, r6, r4
206: eb47 0705 adc.w r7, r7, r5
20a: e9c0 6700 strd r6, r7, [r0]
20e: e9c0 2302 strd r2, r3, [r0, #8]
212: bcf0 pop {r4, r5, r6, r7}
214: b002 add sp, #8
216: 4770 bx lr
fixed up the adds
.thumb_func
.globl test2
test2:
sub sp, #8
push {r4, r5, r6, r7}
ldrd r4, r5, [sp, #24]
add r1, sp, #16
stmia r1, {r2, r3}
ldrd r2, r3, [sp, #40]
add r2, r4
adc r3, r5
ldrd r6, r7, [sp, #16]
ldrd r4, r5, [sp, #32]
adc r6, r4
adc r7, r5
strd r6, r7, [r0]
strd r2, r3, [r0, #8]
pop {r4, r5, r6, r7}
add sp, #8
bx lr
final result
00000024 <test2>:
24: b082 sub sp, #8
26: b4f0 push {r4, r5, r6, r7}
28: e9dd 4506 ldrd r4, r5, [sp, #24]
2c: a904 add r1, sp, #16
2e: c10c stmia r1!, {r2, r3}
30: e9dd 230a ldrd r2, r3, [sp, #40] ; 0x28
34: 1912 adds r2, r2, r4
36: 416b adcs r3, r5
38: e9dd 6704 ldrd r6, r7, [sp, #16]
3c: e9dd 4508 ldrd r4, r5, [sp, #32]
40: 4166 adcs r6, r4
42: 416f adcs r7, r5
44: e9c0 6700 strd r6, r7, [r0]
48: e9c0 2302 strd r2, r3, [r0, #8]
4c: bcf0 pop {r4, r5, r6, r7}
4e: b002 add sp, #8
50: 4770 bx lr
Notice the fewer number of thumb2 instructions, unless you are on a cortex-A that has thumb2 support, those fetches from flash (cortex-m) are (can be) slow. I see you are trying to save the push and pop of two more registers but you cost yourself more fetches. You could take the above and still re-arrange the loads and stores and save those two registers.
minimal testing so far. printfs show the upper words adding, where I didnt see that with your code. I am stilll trying to unwind the calling convention (please document your code more for us), looks like r0 is prepped by the caller to place the result, rest is on the stack. I am using a stellaris launchpad (cortex-m4).

Resources