Hard Fault when calling __main()

lpcware — Thu, 16 Jun 2016 00:35:22 GMT

Content originally posted in LPCWare by krowberr on Thu May 26 16:27:52 MST 2016
I am using the Debug Build Configuration. When I attempt to run the code, I get a Hard Fault in c_startup_lpc1768.c when I make the __main() call. I cam make this go away by renaming my new file from filename.c to filename.c.NotUsed.

The code size is smaller without the file as this is the Debug Build Configuration, unused functions to not appear to be optimized out. That the application links with and without the file shows that the code isn't used anywhere.

I am using the Free Edition of LPCXpresso. Is this what happens when I exceed the allowed Debug code size? It looks like I still have plenty of space available. But the code size increse does seem to break it.

I haven't modified the c_startup_lpc1768.c file at all.

Details on the two images

Works:

From the linker:
   text   data    bss    dec    hexfilename
200112   3572 59428 263112 403c8RockmelonC.axf

At the bottom of the Debug Window:
TargetDownload size = 203684, address=0x170, msg=, rate= 811488

Doesn't work:

From the linker:
   text   data    bss    dec    hexfilename
200704   3572 59436 263712 40620RockmelonC.axf

At the bottom of the Debug WindowL
TargetDowload size = 204276, address=0x170, msg=, rate=841072

Display license type
License type: FULL
Debug limit: 256k
Activation code:…

A FULL copy of LPCXpresso (Free Edition) may be used for production.

Re: Hard Fault when calling __main()

lpcware — Thu, 16 Jun 2016 00:35:23 GMT

Content originally posted in LPCWare by lpcxpresso-support on Fri May 27 05:26:50 MST 2016
So the LPC17xx project wizard in LPCXpresso IDE would normally create a startup file called - cr_startup_lpc175x_6x.c. So as your startup filename doesn't match this, it indicates either a very old version or one that has been obtained from another source?

Anyway, in the first place, I would suggest creating a new LPC1768 project using the New project wizard and checking that this builds and runs (or else try one of the LPC17 LPCOpen examples if you board matches). If these work, then compare with your own project.

You could also try to work out why and where you are getting a hard fault:

https://www.lpcware.com/content/faq/lpcxpresso/debugging-hard-fault

If you want someone to look at your project, then export it and post it in a ZIP file:

https://www.lpcware.com/content/faq/lpcxpresso/how-importexport-projects

And if you want more help then please also confirm the version of LPCXpresso IDE you are using, the debug probe and also the MCU and target board that you have.

Regards,
LPCXpresso Support

Re: Hard Fault when calling __main()

lpcware — Thu, 16 Jun 2016 00:35:24 GMT

Content originally posted in LPCWare by krowberr on Tue May 31 14:49:50 MST 2016
It is an old version of the startup file as this is an existing project. I replaced the old startup file with the new. I still got Hard Faults as __main() was being called.

I am using LPCXpresso v7.9.2 [Build 493] [2015-09-14], the Free License.

From the Debug Console:
LPCXpresso Debug Driver v7.9 (Aug 11 2015 15:47:24 - crt_emu_cm3_nxp build 262)
Emu(1): Connected&Reset. CpuID: 410FC230. Info: FTVCUDKEA
Debug protocol: SWD Frequency 3000 KHz. RTCK: Disabled. Vector catch: Disabled.
loaded v.2 On-chip Flash Memory C:\nxp\LPCXpresso_7.9.2_493\lpcxpresso\bin\Flash\LPC175x_6x_512.cfx
image 'LPC175x_6x (512K) Aug 11 2015 16:53:22'
NXP: LPC1768 Part ID: 0x00000000
Connected: was_reset=true. was_stopped=false
LPCXpresso Free License - Download limit is 256K
Writing 204276 bytes to address 0x00000000 in Flash
Progress meter completed at over 100% (204800/204276 bytes)
Erased/Wrote page 0-20 with 204276 bytes in 1652msec
Flash Write Done
Flash Program Summary: 204276 bytes in 1.65 seconds (120.76 KB/sec)
nSRST assert (if available)
Executing in user flash.

Stopped: VectorCatch:HardF (PC was 0x00000000) (VectorCatch)

Both highlights are mine.

It is quite possible that we have exceeded the Free Version size limit. However, I don't see any obvious messages to that effect. I don't want to buy licenses if that doesn't fix the problem. If that is the problem, where are the tools telling me that the image is too large?

The gdb trace shows:
173,277 2-gdb-version
173,282 ~"GNU gdb (GNU Tools for ARM Embedded Processors) 7.8.0.20150604-cvs\n"
173,282 ~"Copyright (C) 2014 Free Software Foundation, Inc.\n"
173,282 ~"License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>\nThis is fre\
e software: you are free to change and redistribute it.\nThere is NO WARRANTY, to the extent permitt\
ed by law. Type \"show copying\"\nand \"show warranty\" for details.\n"
173,282 ~"This GDB was configured as \"--host=i686-w64-mingw32 --target=arm-none-eabi\".\nType \"sho\
w configuration\" for configuration details."
173,282 ~"\nFor bug reporting instructions, please see:\n"
173,282 ~"<http://www.gnu.org/software/gdb/bugs/>.\n"
173,282 ~"Find the GDB manual and other documentation resources online at:\n<http://www.gnu.org/soft\
ware/gdb/documentation/>.\n"
173,283 ~"For help, type \"help\".\n"
173,283 ~"Type \"apropos word\" to search for commands related to \"word\".\n"
173,283 2^done
173,283 (gdb)
173,284 3-environment-cd C:/ClopeDev/SeedCDMA/Workspace/RockmelonC
173,293 3^done
173,293 (gdb)
173,293 4-gdb-set breakpoint pending on
173,303 4^done
173,303 (gdb)
173,303 5-enable-pretty-printing
173,313 5^done
173,313 (gdb)
173,313 6-gdb-set python print-stack none
173,323 6^done
173,323 (gdb)
173,323 7-gdb-set print object on
173,333 7^done
173,333 (gdb)
173,333 8-gdb-set print sevenbit-strings on
173,343 8^done
173,343 (gdb)
173,343 9-gdb-set charset ISO-8859-1
173,353 9^done
173,353 (gdb)
173,354 10-gdb-set auto-solib-add on
173,363 10^done
173,363 (gdb)
173,370 11-file-exec-and-symbols --thread-group i1 C:/ClopeDev/SeedCDMA/Workspace/RockmelonC/Debug/R\
ockmelonC.axf
173,414 11^done
173,414 (gdb)
173,421 12-interpreter-exec console "set remotetimeout 60000"
173,424 =cmd-param-changed,param="remotetimeout",value="60000"
173,424 12^done
173,424 (gdb)
173,425 13-target-select extended-remote | crt_emu_cm3_nxp -msg-port=59745 -swv-port=59746 -g -mi -2\
-pLPC1768 -vendor=NXP -e1 -flash-driver=LPC175x_6x_512.cfx
174,273 =thread-group-started,id="i1",pid="42000"
174,273 =thread-created,id="1",group-id="i1"
174,277 14-list-thread-groups --available
174,378 ~"0x00000160 in bss_init (start=268439800, len=28192) at ../src/cr_startup_lpc176x.c:245\n"
174,379 ~"245\t for (loop = 0; loop < len; loop = loop + 4)\n"
174,379 *stopped,frame={addr="0x00000160",func="bss_init",args=[{name="start",value="268439800"},{na\
me="len",value="28192"}],file="../src/cr_startup_lpc176x.c",fullname="C:\\ClopeDev\\SeedCDMA\\Worksp\
ace\\RockmelonC\\src\\cr_startup_lpc176x.c",line="245"},thread-id="1",stopped-threads="all"
174,381 13^connected
174,381 (gdb)
174,381 14^error,msg="Can not fetch data now."
174,381 (gdb)
174,382 15-interpreter-exec console "set mem inaccessible-by-default off"
174,383 =cmd-param-changed,param="mem inaccessible-by-default",value="off"
174,383 15^done
174,383 (gdb)
174,384 16-interpreter-exec console "mon ondisconnect cont"
174,395 16^done
174,395 (gdb)
174,396 17-interpreter-exec console "set arm force-mode thumb"
174,396 =cmd-param-changed,param="arm force-mode",value="thumb"
174,396 17^done
174,396 (gdb)
174,397 18-interpreter-exec console "mon capabilities"
174,417 18^done
174,417 (gdb)
174,418 19-interpreter-exec console "set remote hardware-breakpoint-limit 6"
174,419 =cmd-param-changed,param="remote hardware-breakpoint-limit",value="6"
174,419 19^done
174,419 (gdb)
174,420 20-interpreter-exec console "mon semihost enable"
174,429 20^done
174,429 (gdb)
174,431 21-interpreter-exec console "mon info,all"
174,441 21^done
174,441 (gdb)
174,542 22-data-list-register-names
174,547 22^done,register-names=["r0","r1","r2","r3","r4","r5","r6","r7","r8","r9","r10","r11","r12",\
"sp","lr","pc","","","","","","","","","","xpsr","","","","","","","","","","","","","","","","","",\
"","","","","","","","","","","","","","","","","","","","","","","","","","","","","","","","","","\
","","","","","","","","","","","","","","","cycles"]
174,547 (gdb)
174,548 23-target-download
174,577 23+download,{section=".text",section-size="200704",total-size="1141513"}
174,577 23+download,{section=".text",section-sent="16112",section-size="200704",total-sent="16112",t\
otal-size="1141513"}
174,613 24-list-thread-groups
174,717 23+download,{section=".data",section-size="3572",total-size="1141513"}
176,579 23^done,address="0x170",load-size="204276",transfer-rate="820384",write-rate="12767"
176,579 (gdb)
176,580 24^done,groups=[{id="i1",type="process",pid="42000",executable="C:\\ClopeDev\\SeedCDMA\\Work\
space\\RockmelonC\\Debug\\RockmelonC.axf"}]
176,580 (gdb)
176,613 25-gdb-show --thread-group i1 language
176,618 26-list-thread-groups i1
176,620 25^done,value="auto"
176,620 (gdb)
176,620 27-gdb-set --thread-group i1 language c
176,630 26^done,threads=[{id="1",target-id="Thread <main>",frame={level="0",addr="0x00000170",func="\
ResetISR",args=[],file="../src/cr_startup_lpc176x.c",fullname="C:\\ClopeDev\\SeedCDMA\\Workspace\\Ro\
ckmelonC\\src\\cr_startup_lpc176x.c",line="268"},state="stopped"}]
176,630 (gdb)
176,630 27^done
176,630 (gdb)
176,630 28-interpreter-exec --thread-group i1 console "p/x (char)-1"
176,633 29-stack-info-depth --thread 1 11
176,640 ~"$1 = 0xff\n"
176,640 28^done
176,640 (gdb)
176,640 29^done,depth="1"
176,640 (gdb)
176,641 30-data-evaluate-expression --thread-group i1 "sizeof (void*)"
176,641 31-stack-list-frames --thread 1
176,651 30^done,value="4"
176,651 (gdb)
176,651 31^done,stack=[frame={level="0",addr="0x00000170",func="ResetISR",file="../src/cr_startup_lp\
c176x.c",fullname="C:\\ClopeDev\\SeedCDMA\\Workspace\\RockmelonC\\src\\cr_startup_lpc176x.c",line="2\
68"}]
176,651 (gdb)
176,652 32-gdb-set --thread-group i1 language auto
176,661 32^done
176,661 (gdb)
176,661 33-interpreter-exec --thread-group i1 console "show endian"
176,661 ~"The target endianness is set automatically (currently little endian)\n"
176,661 33^done
176,661 (gdb)
176,664 34-break-insert -t -f main
176,712 34^done,bkpt={number="1",type="breakpoint",disp="del",enabled="y",addr="0x00002c84",func="ma\
in",file="../src/main.c",fullname="C:\\ClopeDev\\SeedCDMA\\Workspace\\RockmelonC\\src\\main.c",line=\
"60",thread-groups=["i1"],times="0",original-location="main"}
176,712 (gdb)
176,715 35-exec-continue --thread-group i1
176,715 36-thread-info 1
176,722 ~"Note: automatically using hardware breakpoints for read-only addresses.\n"
176,732 35^running
176,732 *running,thread-id="all"
176,732 (gdb)
176,844 ~"\nProgram received signal "
176,844 ~"SIGSTOP, Stopped (signal).\n"
176,854 ~"HardFault_Handler () at ../src/HardFault_Handler.c:33\n"
176,855 ~"33\t __asm(\n"
176,855 *stopped,reason="signal-received",signal-name="SIGSTOP",signal-meaning="Stopped (signal)",fr\
ame={addr="0x00002a78",func="HardFault_Handler",args=[],file="../src/HardFault_Handler.c",fullname="\
C:\\ClopeDev\\SeedCDMA\\Workspace\\RockmelonC\\src\\HardFault_Handler.c",line="33"},thread-id="1",st\
opped-threads="all"
176,855 (gdb)
176,855 36^done,threads=[{id="1",target-id="Thread <main>",frame={level="0",addr="0x00002a78",func="\
HardFault_Handler",args=[],file="../src/HardFault_Handler.c",fullname="C:\\ClopeDev\\SeedCDMA\\Works\
pace\\RockmelonC\\src\\HardFault_Handler.c",line="33"},state="stopped"}]
176,855 (gdb)
176,943 37-stack-info-depth --thread 1 11
176,985 37^done,depth="3"
176,985 (gdb)
176,988 38-stack-list-frames --thread 1
176,990 39-stack-list-frames --thread 1 0 1
176,995 38^done,stack=[frame={level="0",addr="0x00002a78",func="HardFault_Handler",file="../src/Hard\
Fault_Handler.c",fullname="C:\\ClopeDev\\SeedCDMA\\Workspace\\RockmelonC\\src\\HardFault_Handler.c",\
line="33"},frame={level="1",addr="0xfffffff9",func="<signal handler called>"},frame={level="2",addr=\
"0x00000000",func="g_pfnVectors"}]
176,995 (gdb)
176,995 39^done,stack=[frame={level="0",addr="0x00002a78",func="HardFault_Handler",file="../src/Hard\
Fault_Handler.c",fullname="C:\\ClopeDev\\SeedCDMA\\Workspace\\RockmelonC\\src\\HardFault_Handler.c",\
line="33"},frame={level="1",addr="0xfffffff9",func="<signal handler called>"}]
176,995 (gdb)

Re: Hard Fault when calling __main()

lpcware — Thu, 16 Jun 2016 00:35:24 GMT

Content originally posted in LPCWare by lpcxpresso-support on Tue May 31 16:25:54 MST 2016

It's looks like some code is calling through a null pointer. Although you're not up against the code size limit, something is getting squeezed. Note that __main is linked as a weak external. In some circumstances, it may call the _initio library code. This call, in turn, uses the heap. This means you need a valid heap pointer. Other possibilities include your stack running into your data, your heap (if used) growing into your stack, etc..

Thanks and regards,
LPCXpresso Support

Re: Hard Fault when calling __main()

lpcware — Thu, 16 Jun 2016 00:35:25 GMT

Content originally posted in LPCWare by krowberr on Thu Jun 02 10:53:53 MST 2016
Thanks!! VERY helpful.

I am running out of RAM, not Flash. I reduced the size of an array to confirm and the code runs. I need a long term fix. But you pointed me in the right direction.

This issue has been resolved.

Re: Hard Fault when calling __main()

lpcware — Sun, 19 Jun 2016 01:11:52 GMT

bump

topic Hard Fault when calling __main() in LPCXpresso IDE

Hard Fault when calling __main()

Re: Hard Fault when calling __main()

Re: Hard Fault when calling __main()

Re: Hard Fault when calling __main()

Re: Hard Fault when calling __main()

Re: Hard Fault when calling __main()