Menu
▾
▴
[sdcc-devel] [ sdcc-Bugs-1353379 ] SIGABRT compiling PIC USB example MCHPFSUSB
|
From: SourceForge.net <no...@so...> - 2005-12-08 09:50:46
|
Bugs item #1353379, was opened at 2005-11-10 21:07 Message generated for change (Comment added) made by vrokas You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=100599&aid=1353379&group_id=599 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: pic16 target Group: None Status: Open Resolution: None Priority: 5 Submitted By: Ezra Buehler (easyb) Assigned to: Raphael Neider (tecodev) Summary: SIGABRT compiling PIC USB example MCHPFSUSB Initial Comment: SDCC crashes when trying to build the by julian Green ported USB example code from Microchip. Easys-Mac:/Volumes/Users/easyb/Prog/PIC/sdcc/MCHPFSUSB/ fw/Cdc easyb$ make sdcc -mpic16 -p18f2550 -I . -c -c system/usb/usbdrv/usbdrv.c -o system/usb/usbdrv/usbdrv.o Processor: 18f2550 sdcc -mpic16 -p18f2550 -I . -c -c system/usb/usbmmap.c -o system/usb/usbmmap.o Processor: 18f2550 sdcc -mpic16 -p18f2550 -I . -c -c main.c -o main.o Processor: 18f2550 sdcc -mpic16 -p18f2550 -I . -c -c system/usb/usb9/usb9.c -o system/usb/usb9/usb9.o Processor: 18f2550 gen.c:10532: failed assertion `IS_DATA_PTR(operandType(left))' Caught signal 6: SIGABRT make: *** [system/usb/usb9/usb9.o] Error 1 Easys-Mac:/Volumes/Users/easyb/Prog/PIC/sdcc/MCHPFSUSB/ fw/Cdc easyb$ I use the latest sdcc snapshot for Mac OS X. SDCC : mcs51/gbz80/z80/avr/ds390/pic16/pic14/TININative/xa51/ ds400/hc08 2.5.4 #1151 (Nov 10 2005) (UNIX) Following has been reported for the win32 snapshot: There are some compiling problems with the following three C files under SDCC snapshot. system/usb/usb9/usb9.c system/usb/usbctrltrf/usbctrltrf.c system/usb/class/cdc/cdc.c Error message: Assertion failed: IS_DATA_PTR(operandType(left)), file gen.c, line 10532 The source code can be found at: http://electro.easyb.ch/files/MCHPFSUSB.tgz It has been reported to compile with 2.5.3 Oct 19 2005 and 2.5.3 Sep 19 2005 snapshots If you need more information/help don't hesitate to mail me. easy- bATfreesurf.ch Thanx in advance, Ezra. ---------------------------------------------------------------------- >Comment By: Vangelis Rokas (vrokas) Date: 2005-12-08 11:50 Message: Logged In: YES user_id=770505 Raphael, this is a bug in the gen.c. Assertion is placed there to hunt a possible future bug. I have the solution for this and I'll commit it as soon as possible. Vangelis ---------------------------------------------------------------------- Comment By: Raphael Neider (tecodev) Date: 2005-12-07 23:43 Message: Logged In: YES user_id=1115835 Hi Ezra, Sorry for the delay: The uploaded file does not help me very much... What I would really need is output from a debugger: a stack trace (e.g. gdb --args sdcc -mpic16 -p18f2550 -I . -c -c system/usb/usb9/usb9.c -o system/usb/usb9/usb9.o [Enter] run [Enter]; upon segfault: bt [Enter] This should bring up a list of functions that have been called along with source line numbers. I hope, this works similarly on Mac OS...). It would be great if you could supply something like that... Regards, Raphael Neider ---------------------------------------------------------------------- Comment By: Ezra Buehler (easyb) Date: 2005-11-19 14:44 Message: Logged In: YES user_id=559151 Sorry about that. I now uploaded the file. Hope it's any good. Otherwise tell me what to do and I'll try. Unfortunately I will not be able to use my mac for the next two weeks. Except on the weekends. But thanx for your help. Cheers, Ezra. ---------------------------------------------------------------------- Comment By: Raphael Neider (tecodev) Date: 2005-11-16 16:28 Message: Logged In: YES user_id=1115835 I cannot reproduce the SIGSEGV on my Linux box. Your crash log seems not to have it made into the tracker. Please upload or mail it to me. A gdb backtrace might be helpful, but is possibly included in Mac OS's crash log. BTW: The processor used during compilation only defines the memory layout; code generation itself is not influenced. Regards, Raphael ---------------------------------------------------------------------- Comment By: Ezra Buehler (easyb) Date: 2005-11-12 19:07 Message: Logged In: YES user_id=559151 Ok, tried the latest Mac OS snapshot. No SIGABRT now, but sdcc crashes with a SIGSEGV instead. sdcc -mpic16 -p18f2550 -I . -c -c system/usb/usbctrltrf/usbctrltrf.c -o system/usb/usbctrltrf/usbctrltrf.o Processor: 18f2550 Caught signal 11: SIGSEGV make: *** [system/usb/usbctrltrf/usbctrltrf.o] Error 1 I noticed that the MC2.tbz uses the 18f4455, does that make such a difference? I also attached the Mac OS X crash log, maybe it helps. Cheers, Ezra. ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2005-11-12 10:09 Message: Logged In: NO Sorry for the confusion. The only issue is: gen.c:10532: failed assertion `IS_DATA_PTR(operandType(left))' Caught signal 6: SIGABRT I've been told that compiling with sdcc using the -c option always worked, but you had to link manually. Anyway I didn't write the code and have to admit that I didn't really look at it yet. But John Swensen did and somehow made it build fine with the current snapshots: http://swenmac.homeip.net:8080/MC2.tbz So I'll wait for the 1152 snapshot and see. Thanx a lot, Ezra. ---------------------------------------------------------------------- Comment By: Raphael Neider (tecodev) Date: 2005-11-12 01:15 Message: Logged In: YES user_id=1115835 Fixed in SDCC 2.5.4 build #1152. system/usb/usb9/usb9.c: with `compiling problems' do you mean the warning "pcoderegs.c:367: removing reg r0x00 because it is used only once"? This is harmless and scheduled for removal... system/usb/usbctrltrf/usbctrltrf.c, system/usb/class/cdc/cdc.c: I cannot make out any problems here---did they simply fail because of the assert() statement or do you experience a different problem? Raphael ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=100599&aid=1353379&group_id=599 |