From nathan at codesourcery.com Fri Dec 1 08:20:15 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Fri, 01 Dec 2006 08:20:15 +0000 Subject: Update release Message-ID: <456FE5BF.4050206@codesourcery.com> People may be wondering why there's been no update and it is now December. P&E micro have provided an updated driver for their wigglers and we are incorporating that into our toolchains. It fixes a problem we had encountered, and it is better to wait a few days while we test it. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From fred at aidounix.com Tue Dec 5 12:45:21 2006 From: fred at aidounix.com (fred at aidounix.com) Date: Tue, 5 Dec 2006 13:45:21 +0100 (CET) Subject: Toolchain, asm parameters numbering and optimisation flags Message-ID: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> Hello, I'm just trying to optimize the speex codec for the coldfire CPU. Using the CodeSourcery toolchain, I have strange behaviour (I think) of GCC. I've attached the C source code, then the results of : m68k-uclinux-gcc -Wa,--register-prefix-optional -Wa,-memac -Wa,-mcpu=5208 -mcpu=5208 -O2 -S -o test.O2.s test.c and m68k-uclinux-gcc -Wa,--register-prefix-optional -Wa,-memac -Wa,-mcpu=5208 -mcpu=5208 -O0 -S -o test.O0.s test.c For the O2 flag, only the D0 register is used, mixing it as a temporary data holder, and as a results holder. For the O0 flag, as you can see line 37 of test.O0.s it is used as the "len" parameter and as the "sum" result variable. I'm really lost. I have only ARM inline asm experience. I suspect it's a very simple problem, but I can't, after days of search, find a solution nor an explication Does someone hav an hint? please :-) Thanx in advance -------------- next part -------------- A non-text attachment was scrubbed... Name: test.c Type: text/x-csrc Size: 1196 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: test.O0.s Type: application/octet-stream Size: 1692 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: test.O2.s Type: application/octet-stream Size: 1211 bytes Desc: not available URL: From fred at aidounix.com Tue Dec 5 12:52:44 2006 From: fred at aidounix.com (fred at aidounix.com) Date: Tue, 5 Dec 2006 13:52:44 +0100 (CET) Subject: addendum Message-ID: <37933.82.233.75.208.1165323164.squirrel@mail.aidounix.com> Oups, I forgot to write down my gcc version : m68k-uclinux-gcc (GCC) 4.1.0 (CodeSourcery Freescale 4.1-14) TIA From nathan at codesourcery.com Tue Dec 5 13:18:44 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 05 Dec 2006 13:18:44 +0000 Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> Message-ID: <457571B4.4080009@codesourcery.com> fred at aidounix.com wrote: > Hello, > > I'm just trying to optimize the speex codec for the coldfire CPU. > Using the CodeSourcery toolchain, I have strange behaviour (I think) > of GCC. > I've attached the C source code, then the results of : > m68k-uclinux-gcc -Wa,--register-prefix-optional -Wa,-memac -Wa,-mcpu=5208 > -mcpu=5208 -O2 -S -o test.O2.s test.c > 35 : "=d" (sum) > 36 : "a"(x), "a"(y), "d"(len), "d"(t1), "d"(t2), "d"(t3) > 37 :"cc" Here's your problem. The inline asm semantics are that all the input operands are read before any of the output operands are written. thus it is safe to allocate the same register(s) for inputs and outputs. All gcc targets are the same in this regard. Your assembly breaks that. you can fix this by telling gcc that 'sum' is an early clobber :"=&d" (sum) I notice that you are also modifying the input operands. That's breaking the semantics too. the compiler will presume the register it placed 'len' into before the asm has the same value after the asm. IIRC you indicate this altering of input values by making them output operands, but using a + rather than an =. oh, you don't need the -Wa,-memac -Wa,-mcpu=5208 flags. The latter flag implies the former, and the compiler's -mcpu=5208 option is passed through to the assembler anyway (besides telling the compiler that you're targetting a 5208). nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From fred at aidounix.com Tue Dec 5 23:14:57 2006 From: fred at aidounix.com (fred at aidounix.com) Date: Wed, 6 Dec 2006 00:14:57 +0100 (CET) Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <457571B4.4080009@codesourcery.com> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> Message-ID: <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> > Here's your problem. The inline asm semantics are that all the input > operands > are read before any of the output operands are written. thus it is safe > to > allocate the same register(s) for inputs and outputs. All gcc targets are > the > same in this regard. Your assembly breaks that. you can fix this by > telling > gcc that 'sum' is an early clobber > :"=&d" (sum) > Hi Nathan, your knowledge is some order of magnitude beyound mine! :) I've corrected those points, now I can compile with -O0. But GCC still scramble the registers (assigning D0 to %2 and %3 for example). for example : C source : move.l (%[X])+, %[tmp1]; move.l (%[Y])+, %[tmp2]; move.w %[tmp1], %[R1]; move.w %[tmp2], %[R2]; generates, with -O2 : move.l (a0)+, d1; move.l (a1)+, d1; ouch! move.w d1, d2; move.w d1, d3; I really don't understand why it nicelly works with -O0 and not with -O2. Any idea? Thanx again From dan at codesourcery.com Tue Dec 5 23:14:39 2006 From: dan at codesourcery.com (Daniel Jacobowitz) Date: Tue, 5 Dec 2006 18:14:39 -0500 Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> Message-ID: <20061205231438.GO30862@caradoc.them.org> On Wed, Dec 06, 2006 at 12:14:57AM +0100, fred at aidounix.com wrote: > for example : > C source : > move.l (%[X])+, %[tmp1]; > move.l (%[Y])+, %[tmp2]; What are the constraints for tmp1 and tmp2? > I really don't understand why it nicelly works with -O0 and not > with -O2. You're probably just getting luckier at -O0. -- Daniel Jacobowitz CodeSourcery From dan at codesourcery.com Tue Dec 5 23:54:39 2006 From: dan at codesourcery.com (Daniel Jacobowitz) Date: Tue, 5 Dec 2006 18:54:39 -0500 Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <57018.82.233.75.208.1165361042.squirrel@mail.aidounix.com> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> <20061205231438.GO30862@caradoc.them.org> <57018.82.233.75.208.1165361042.squirrel@mail.aidounix.com> Message-ID: <20061205235438.GP30862@caradoc.them.org> This is the same problem as Nathan was talking about. On Wed, Dec 06, 2006 at 12:24:02AM +0100, fred at aidounix.com wrote: > "move.l (%[X])+, %[e1];\n" > "move.l (%[Y])+, %[e2];\n" > "swap %[e1];" > "swap %[e2];" > "move.w %[e1], %[t1];" > "move.w %[e2], %[t2];" You are setting these. Therefore they are not inputs to your asm; inputs are only read, not modified. > : [sum]"=&d" (sum), [len]"+&d"(len) > : [X]"a"(x), [Y]"a"(y), [t1]"d"(t1), [t2]"d"(t2), [e1]"d"(e1), [e2]"d"(e2) But this says that t1 is only an input. -- Daniel Jacobowitz CodeSourcery From fred at aidounix.com Wed Dec 6 00:58:22 2006 From: fred at aidounix.com (fred at aidounix.com) Date: Wed, 6 Dec 2006 01:58:22 +0100 (CET) Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <20061205231438.GO30862@caradoc.them.org> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> <20061205231438.GO30862@caradoc.them.org> Message-ID: <57015.82.233.75.208.1165366702.squirrel@mail.aidounix.com> Thanx a lot Daniel & Nathan. It does compile with -O2 now. So, I can start my part... ColdFire optimisations.... :-p Could you, please, redirect me to a tutorial on the art of writing constraints? Have a nice night From dan at codesourcery.com Wed Dec 6 01:18:51 2006 From: dan at codesourcery.com (Daniel Jacobowitz) Date: Tue, 5 Dec 2006 20:18:51 -0500 Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <57015.82.233.75.208.1165366702.squirrel@mail.aidounix.com> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> <20061205231438.GO30862@caradoc.them.org> <57015.82.233.75.208.1165366702.squirrel@mail.aidounix.com> Message-ID: <20061206011850.GR30862@caradoc.them.org> On Wed, Dec 06, 2006 at 01:58:22AM +0100, fred at aidounix.com wrote: > Could you, please, redirect me to a tutorial on the art of writing > constraints? I don't know of one, sorry - there's a section in the GCC manual about it, but it's not easy to understand. -- Daniel Jacobowitz CodeSourcery From nathan at codesourcery.com Wed Dec 6 06:58:38 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Wed, 06 Dec 2006 06:58:38 +0000 Subject: [coldfire-gnu-discuss] Toolchain, asm parameters numbering and optimisation flags In-Reply-To: <20061206011850.GR30862@caradoc.them.org> References: <50071.82.233.75.208.1165322721.squirrel@mail.aidounix.com> <457571B4.4080009@codesourcery.com> <57150.82.233.75.208.1165360497.squirrel@mail.aidounix.com> <20061205231438.GO30862@caradoc.them.org> <57015.82.233.75.208.1165366702.squirrel@mail.aidounix.com> <20061206011850.GR30862@caradoc.them.org> Message-ID: <45766A1E.1020303@codesourcery.com> Daniel Jacobowitz wrote: > On Wed, Dec 06, 2006 at 01:58:22AM +0100, fred at aidounix.com wrote: >> Could you, please, redirect me to a tutorial on the art of writing >> constraints? > > I don't know of one, sorry - there's a section in the GCC manual about > it, but it's not easy to understand. Yes, unfortunately there isn't one. the compiler internals manual does describe it, but that is not really in tutorial form. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From mark at codesourcery.com Fri Dec 8 19:46:47 2006 From: mark at codesourcery.com (Mark Mitchell) Date: Fri, 08 Dec 2006 11:46:47 -0800 Subject: Sourcery G++ 4.1-30 for ColdFire now available Message-ID: <4579C127.6070500@codesourcery.com> CodeSourcery is pleased to announce a new version of Sourcery G++ for ColdFire. Highlights of this release include: * Support for Command Converter Server (CCS) debugging * Improved support for hardware breakpoints when running programs from flash memory * Semihosting no longer uses trap 15, so all interrupt vectors are available to the target program * Newer versions of GDB and Binutils are included. * P&E hardware drivers are included * Much more! Visit: http://www.codesourcery.com/gnu_toolchains/coldfire/ to download this new release of Sourcery G++. Enjoy! -- Mark Mitchell CodeSourcery mark at codesourcery.com (650) 331-3385 x713 From Stephen.Pickering at blueyonder.co.uk Sat Dec 9 11:22:24 2006 From: Stephen.Pickering at blueyonder.co.uk (Stephen Pickering) Date: Sat, 09 Dec 2006 11:22:24 +0000 Subject: [coldfire-gnu-discuss] Sourcery G++ 4.1-30 for ColdFire now available Message-ID: <457A9C70.9090609@blueyonder.co.uk> I've installed it on Mandriva 2007. Now how do I connect to a Coldfire board using the P&E usb bdm ? Version 4.1.14 had a p&e stub (didn't work), latest version has something called m68k-elf-sprite: $ m68k-elf-sprite -v -i m68k-elf-sprite:CodeSourcery ColdFire Debug Sprite (CodeSourcery Sourcery G++ 4.1-30) CodeSourcery ColdFire Debug Sprite (CodeSourcery Sourcery G++ 4.1-30) ccs: [timeout=;speed=] CCS Adaptor ccs://$Host:$Port/$Chainpos - CCS address pe: [speed=;memory-timeout=] P&E Adaptor pe: P&E DLL not present I have P&E linux driver installed: pe_driver.ver.318.2.802 and bdm is detected: # lsusb -v Bus 007 Device 002: ID 1357:0503 Device Descriptor: bLength 18 bDescriptorType 1 bcdUSB 2.00 bDeviceClass 0 (Defined at Interface level) bDeviceSubClass 0 bDeviceProtocol 0 bMaxPacketSize0 64 idVendor 0x1357 idProduct 0x0503 bcdDevice 2.00 iManufacturer 1 iProduct 2 USB-ML-CF Rev C iSerial 3 PE6011459 bNumConfigurations 1 Any help? -- Steve bcdDevice 2.00 -------------- next part -------------- An HTML attachment was scrubbed... URL: From claudesylvain at sympatico.ca Sat Dec 9 14:21:21 2006 From: claudesylvain at sympatico.ca (Claude Sylvain) Date: Sat, 09 Dec 2006 09:21:21 -0500 Subject: Connecting to a Coldfire target using P&E USB Coldfire Multilink. Message-ID: <7.0.1.0.2.20061209090600.0232daf8@sympatico.ca> Hello Steve, Maybe the following can help you. I installed the Coldfire GNU toolchain on a machine that run under Windows XP professionnal SP2. I can connect to a M52235EVB using P&E USB Coldfire Multilink using the following operations: 1) From the command line, execute GDB: m68k-elf-gdb 2) From GDB, type: target remote | m68k-elf-sprite pe://USBMultilink m52235evb 3) From GDB, type: load Remark: GDB seems to have a lot of difficulty when executing the "load" command. Most of the time the command will fail. It seem that waiting some time before doing a "load" command can help. Maybe there are some parameters that can be set in "sprite" that can resolve this issue. Claude. From csylvain at electro-technica.com Sat Dec 9 19:34:03 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sat, 09 Dec 2006 14:34:03 -0500 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. Message-ID: <7.0.1.0.2.20061209142633.02329ba0@sympatico.ca> > Linux doesn't use DLL's, does it? No, there is no DLL in Linux. Read the first paragraph on page 25 of "getting-started.pdf" document. That document is included with the tool. You will find all the step to build the P&E driver (a loadable kernel module). Claude From csylvain at electro-technica.com Sat Dec 9 20:15:13 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sat, 09 Dec 2006 15:15:13 -0500 Subject: GCC -I option seems to not work !? Message-ID: <7.0.1.0.2.20061209150513.0231a120@sympatico.ca> Hello, I have installed sourcery G++ Lite, ColdFire ELF 4.1-30 on a computer that run under Window XP professionnal SP2. I am trying to use the -I option when calling GCC (m68k-elf-gcc), but GCC seems to ignore the option. I have tryed the following, without success: -I -I/ -I./ -I.// -I\ -I.\ -I.\\ Am I missing something ? Claude. From dan at codesourcery.com Sat Dec 9 20:47:16 2006 From: dan at codesourcery.com (Daniel Jacobowitz) Date: Sat, 9 Dec 2006 15:47:16 -0500 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? In-Reply-To: <7.0.1.0.2.20061209150513.0231a120@sympatico.ca> References: <7.0.1.0.2.20061209150513.0231a120@sympatico.ca> Message-ID: <20061209204715.GF30862@caradoc.them.org> On Sat, Dec 09, 2006 at 03:15:13PM -0500, Claude Sylvain wrote: > Hello, > > I have installed sourcery G++ Lite, ColdFire ELF 4.1-30 on a computer > that run under Window XP professionnal SP2. > > I am trying to use the -I option when calling GCC (m68k-elf-gcc), but > GCC seems to ignore the option. > > I have tryed the following, without success: > > -I > -I/ > -I./ > -I.// > -I\ > -I.\ > -I.\\ > > > Am I missing something ? I don't know, but we need more information to find out. What does the directory structure look like? What's the complete command line you are using to invoke GCC? And are you running it from cmd.exe or from a Cygwin shell? -- Daniel Jacobowitz CodeSourcery From csylvain at electro-technica.com Sun Dec 10 03:35:51 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sat, 09 Dec 2006 22:35:51 -0500 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? Message-ID: <7.0.1.0.2.20061209223406.02343930@sympatico.ca> Hello Daniel, > > Daniel Jacobowitz wrote: > > I don't know, but we need more information to find out. What does the > directory structure look like? What's the complete command line you > are using to invoke GCC? And are you running it from cmd.exe or from a > Cygwin shell? > Here is the directory structure: ----- project \---obj \---out \---src \---drivers | \---ethernet \---startup \---support ----- I use a Makefile to build the project. The Makefile is located in "project/" folder, and executed from that folder. Here is some revealing part of that Makefile: ----- CC = m68k-elf-gcc CFLAGS_COMMON = -mcpu=52235 -Tm52235evb-ram-hosted.ld -I./drivers/ -I./support/ -g -o PSRC = ./src/ PSRC_DRV = ./src/drivers/ POBJ = ./obj/ POUT = ./out/ $(POBJ)main.o: \ $(PSRC)main.c $(PSRC)main.h $(PSRC)project.h $(CC) -c $(CFLAGS_COMMON) $(POBJ)main.o $(PSRC)main.c ----- I run the Makefile from a cmd.exe black box, and typing the following command: cs-make Claude. From mark at codesourcery.com Sun Dec 10 03:58:29 2006 From: mark at codesourcery.com (Mark Mitchell) Date: Sat, 09 Dec 2006 19:58:29 -0800 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? In-Reply-To: <7.0.1.0.2.20061209223406.02343930@sympatico.ca> References: <7.0.1.0.2.20061209223406.02343930@sympatico.ca> Message-ID: <457B85E5.1000706@codesourcery.com> Claude Sylvain wrote: > CFLAGS_COMMON = -mcpu=52235 -Tm52235evb-ram-hosted.ld -I./drivers/ > -I./support/ -g -o The Windows C runtime library does not recognize directory names when they are given with a trailing slash. I think that if you do: -I ./drivers -I ./support it will work. Does that help? -- Mark Mitchell CodeSourcery mark at codesourcery.com (650) 331-3385 x713 From csylvain at electro-technica.com Sun Dec 10 04:06:32 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sat, 09 Dec 2006 23:06:32 -0500 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. Message-ID: <7.0.1.0.2.20061209225809.02262f30@sympatico.ca> Replying to myself: > > Claude wrote: > > GDB seems to have a lot of difficulty when executing the "load" command. > Most of the time the command will fail. It seem that waiting some time > before doing a "load" command can help. Maybe there are some parameters > that can be set in "sprite" that can resolve this issue. > After restarting the computer, the GDB load command work correctly now. Having installed the P&E driver from the disk that come with M52235EVB box, I was thinking that it was not necessary to restart the computer. Claude. From csylvain at electro-technica.com Sun Dec 10 04:24:26 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sat, 09 Dec 2006 23:24:26 -0500 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? Message-ID: <7.0.1.0.2.20061209232247.0230dd68@sympatico.ca> Hello Mark, >> >> Claude Sylvain wrote: >> >> >> CFLAGS_COMMON = -mcpu=52235 -Tm52235evb-ram-hosted.ld -I./drivers/ >> -I./support/ -g -o >> > > Mark Mitchell wrote: > > The Windows C runtime library does not recognize directory names when > they are given with a trailing slash. I think that if you do: > > -I ./drivers -I ./support > > it will work. Does that help? > I have tryed the following: -I./drivers -I./support -I ./drivers -I ./support The problem is still there. GCC seems to not include those directories in its search path. Ok. The current project is rather large. Tomorow, I will start a new basic project that use the -I option. Maybe I will find something. Claude. From dan at codesourcery.com Sun Dec 10 06:10:14 2006 From: dan at codesourcery.com (Daniel Jacobowitz) Date: Sun, 10 Dec 2006 01:10:14 -0500 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? In-Reply-To: <7.0.1.0.2.20061209223406.02343930@sympatico.ca> References: <7.0.1.0.2.20061209223406.02343930@sympatico.ca> Message-ID: <20061210061011.GG30862@caradoc.them.org> On Sat, Dec 09, 2006 at 10:35:51PM -0500, Claude Sylvain wrote: > project > \---obj > \---out > \---src > \---drivers > | \---ethernet > \---startup > \---support Paths for -I are not relative to the source file. They're relative to the current directory. If you need to say "./src/main.c", then you need to say "-I./src/drivers". Try adding the -v option to CFLAGS to see what paths GCC is searching. -- Daniel Jacobowitz CodeSourcery From pop3 at pickering-email.co.uk Sat Dec 9 18:40:18 2006 From: pop3 at pickering-email.co.uk (pickering-email) Date: Sat, 09 Dec 2006 18:40:18 +0000 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <7.0.1.0.2.20061209090600.0232daf8@sympatico.ca> References: <7.0.1.0.2.20061209090600.0232daf8@sympatico.ca> Message-ID: <457B0312.6090805@pickering-email.co.uk> Claude Sylvain wrote: > Hello Steve, > > Maybe the following can help you. > > I installed the Coldfire GNU toolchain on a machine that run under > Windows XP professionnal SP2. I'm using Linux, but here goes... > I can connect to a M52235EVB using P&E USB Coldfire Multilink using > the following operations: > > 1) From the command line, execute GDB: > m68k-elf-gdb $ m68k-elf-gdb GNU gdb (CodeSourcery Sourcery G++ 4.1-30) 6.5.50.20061105-cvs Copyright (C) 2006 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "--host=i686-pc-linux-gnu --target=m68k-elf". For bug reporting instructions, please see: . (gdb) So far, so good. > 2) From GDB, type: > target remote | m68k-elf-sprite pe://USBMultilink m52235evb (gdb) target remote | m68k-elf-sprite pe://USBMultilink m52235evb Remote debugging using | m68k-elf-sprite pe://USBMultilink m52235evb m68k-elf-sprite:error:Cannot load P&E DLL Remote communication error: Connection reset by peer. (gdb) Linux doesn't use DLL's, does it? > 3) From GDB, type: > load > > > Remark: > GDB seems to have a lot of difficulty when executing the "load" > command. Most of the time the command will fail. It seem that > waiting some time before doing a "load" command can help. > Maybe there are some parameters that can be set in "sprite" that can > resolve this issue. > > Claude. > > > > > From csylvain at electro-technica.com Sun Dec 10 15:46:37 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Sun, 10 Dec 2006 10:46:37 -0500 Subject: [coldfire-gnu-discuss] GCC -I option seems to not work !? Message-ID: <7.0.1.0.2.20061210104549.0234adb0@sympatico.ca> Hello Daniel, > Daniel Jacobowitz wrote: > > Paths for -I are not relative to the source file. They're relative to > the current directory. If you need to say "./src/main.c", then you > need to say "-I./src/drivers". > > Try adding the -v option to CFLAGS to see what paths GCC is searching. > You're absolutly right. I made the correction, and all is working fine now. Remark: Yet another stupid error :-| Thank you for your help. Claude. From Rainer at waggerla.de Mon Dec 11 00:46:25 2006 From: Rainer at waggerla.de (Rainer Hauser) Date: Mon, 11 Dec 2006 01:46:25 +0100 Subject: CodeSourcery 4.1-30 Message-ID: <457CAA61.7030405@waggerla.de> i got a error when i try to compile this file (spurios.c): //--------------------------------------------------------- unsigned long spurious = 0; unsigned long intc_spurious = 0; __attribute__ ((interrupt_handler)) void spuriousISR(void) { intc_spurious++; } __attribute__ ((interrupt_handler)) void intc_spuriousISR(void) { spurious++; } //--------------------------------------------------------- m68k-elf-gcc -IG:\Projekte\GPP_Test -IG:\Projekte\GPP_Test\include -IG:\Projekte\GPP_Test\sys -O0 -g -Wall -c -fmessage-length=0 -mcpu=5213 -fomit-frame-pointer -osys\spurious.o ..\sys\spurious.c ..\sys\spurious.c: In function 'spuriousISR': ..\sys\spurious.c:9: error: unable to find a register to spill in class 'GENERAL_REGS' ..\sys\spurious.c:9: error: this is the insn: (insn 8 7 9 1 ..\sys\spurious.c:8 (set (reg:SI 31 [ intc_spurious.0 ]) (mem/c/i:SI (symbol_ref:SI ("intc_spurious") [flags 0x2] ) [0 intc_spurious+0 S4 A16])) 33 {*movsi_cf} (nil) (nil)) ..\sys\spurious.c:9: confused by earlier errors, bailing out any ideas? From nathan at codesourcery.com Mon Dec 11 10:12:25 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Mon, 11 Dec 2006 10:12:25 +0000 Subject: [coldfire-gnu-discuss] Sourcery G++ 4.1-30 for ColdFire now available In-Reply-To: <457A9C70.9090609@blueyonder.co.uk> References: <457A9C70.9090609@blueyonder.co.uk> Message-ID: <457D2F09.4070003@codesourcery.com> Stephen Pickering wrote: > I've installed it on Mandriva 2007. > > Now how do I connect to a Coldfire board using the P&E usb bdm ? > > Version 4.1.14 had a p&e stub (didn't work), latest version has > something called m68k-elf-sprite: > > $ m68k-elf-sprite -v -i > m68k-elf-sprite:CodeSourcery ColdFire Debug Sprite (CodeSourcery > Sourcery G++ 4.1-30) > CodeSourcery ColdFire Debug Sprite (CodeSourcery Sourcery G++ 4.1-30) > ccs: [timeout=;speed=] CCS Adaptor > ccs://$Host:$Port/$Chainpos - CCS address > pe: [speed=;memory-timeout=] P&E Adaptor > pe: P&E DLL not present > > I have P&E linux driver installed: pe_driver.ver.318.2.802 and bdm is > detected: The error message indicates that the PE& library is not installed, or inaccessible. To track this down, use strace strace m68k-elf-sprite -i you should see an open call for libUnit_cfz.so. This is installed in /usr/lib open("/usr/lib/libUnit_cfz.so", O_RDONLY) = 3 nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From nathan at codesourcery.com Mon Dec 11 10:15:00 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Mon, 11 Dec 2006 10:15:00 +0000 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <457B0312.6090805@pickering-email.co.uk> References: <7.0.1.0.2.20061209090600.0232daf8@sympatico.ca> <457B0312.6090805@pickering-email.co.uk> Message-ID: <457D2FA4.2040507@codesourcery.com> pickering-email wrote: > Linux doesn't use DLL's, does it? It has shared objects that can be dlopened, which are the same thing. The error message is confusing though -- we'll fix that. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From nathan at codesourcery.com Mon Dec 11 10:19:20 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Mon, 11 Dec 2006 10:19:20 +0000 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <7.0.1.0.2.20061209225809.02262f30@sympatico.ca> References: <7.0.1.0.2.20061209225809.02262f30@sympatico.ca> Message-ID: <457D30A8.9090301@codesourcery.com> Claude Sylvain wrote: > After restarting the computer, the GDB load command work correctly now. > > Having installed the P&E driver from the disk that come with M52235EVB > box, I > was thinking that it was not necessary to restart the computer. I'm glad you've resolved the issue. Just to clarify though. 1) you had already installed P&E drivers before installing SourceryG++ Lite? 2) Did the SourceryG++ install also install P&E drivers (you'd've got a popup window about that, if it did) 3) Did the P&E driver install (either #1 or #2) request a restart? nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From nathan at codesourcery.com Mon Dec 11 10:56:09 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Mon, 11 Dec 2006 10:56:09 +0000 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 In-Reply-To: <457CAA61.7030405@waggerla.de> References: <457CAA61.7030405@waggerla.de> Message-ID: <457D3949.6010508@codesourcery.com> Rainer Hauser wrote: > i got a error when i try to compile this file (spurios.c): > m68k-elf-gcc -IG:\Projekte\GPP_Test -IG:\Projekte\GPP_Test\include > -IG:\Projekte\GPP_Test\sys -O0 -g -Wall -c -fmessage-length=0 -mcpu=5213 > -fomit-frame-pointer -osys\spurious.o ..\sys\spurious.c > ..\sys\spurious.c: In function 'spuriousISR': > ..\sys\spurious.c:9: error: unable to find a register to spill in class > 'GENERAL_REGS' > ..\sys\spurious.c:9: error: this is the insn: > (insn 8 7 9 1 ..\sys\spurious.c:8 (set (reg:SI 31 [ intc_spurious.0 ]) > (mem/c/i:SI (symbol_ref:SI ("intc_spurious") [flags 0x2] > ) [0 intc_spurious+0 S4 A16])) 33 > {*movsi_cf} (nil) > (nil)) > ..\sys\spurious.c:9: confused by earlier errors, bailing out I can reproduce this problem, and have logged it in our tracker. It'll be fixed in our next public release (date unknown). nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From csylvain at electro-technica.com Mon Dec 11 16:57:21 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Mon, 11 Dec 2006 11:57:21 -0500 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 Message-ID: <7.0.1.0.2.20061211115629.02195db0@electro-technica.com> Hello Nathan, >> >> Rainer Hauser wrote: >> i got a error when i try to compile this file (spurios.c): >> >> >> m68k-elf-gcc -IG:\Projekte\GPP_Test -IG:\Projekte\GPP_Test\include >> -IG:\Projekte\GPP_Test\sys -O0 -g -Wall -c -fmessage-length=0 -mcpu=5213 >> -fomit-frame-pointer -osys\spurious.o ..\sys\spurious.c >> ..\sys\spurious.c: In function 'spuriousISR': >> ..\sys\spurious.c:9: error: unable to find a register to spill in class >> 'GENERAL_REGS' >> ..\sys\spurious.c:9: error: this is the insn: >> (insn 8 7 9 1 ..\sys\spurious.c:8 (set (reg:SI 31 [ intc_spurious.0 ]) >> (mem/c/i:SI (symbol_ref:SI ("intc_spurious") [flags 0x2] >> ) [0 intc_spurious+0 S4 A16])) 33 >> {*movsi_cf} (nil) >> (nil)) >> ..\sys\spurious.c:9: confused by earlier errors, bailing out >> >> > > Nathan wrote: > > > I can reproduce this problem, and have logged it in our tracker. It'll > be fixed in our next public release (date unknown). > - I have the same problem too. - Is there a way to get rid of that problem, or do I have to use the previous version of Sourcery G++ Lite ? Claude. From csylvain at electro-technica.com Mon Dec 11 16:47:35 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Mon, 11 Dec 2006 11:47:35 -0500 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. Message-ID: <7.0.1.0.2.20061211114322.021c22a0@electro-technica.com> Hello Nathan, > > Nathan wrote: > > > 1) you had already installed P&E drivers before installing SourceryG++ Lite? > > 2) Did the SourceryG++ install also install P&E drivers (you'd've got a > popup window about that, if it did) > > 3) Did the P&E driver install (either #1 or #2) request a restart? > - I have installed Sourcery G++ Lite first. - During the installation, I did not see any popup window telling that P&E driver was installed. Maybe, I was looking away at that moment !? - There was no request to restart the computer, and I did not restart it. - Not seeing anything telling that P&E driver was installed, I decide to install it from the supplyed P&E disk that come with the M52235EVB box. - There was no request to restart the computer, and I did not restart it. - Once this driver was installed, I connected the USB Multilink for ColdFire to the computer, and have followed the Windows New Hardware Detected wizard. - There was no request to restart the computer, and I did not restart it. - From that point, I compiled a little project and connected to the target using GDB. From GDB, when executing a "load" command, there was error(s), and I have to make many "load" to finally have the command work correctly. I tryed also to exit GDB, re-execute GDB. Also disconnect/re-connect the USB Multilink without success. - After I have done a Poweroff and Poweron (the following day), all was working like a charm ! Remark: - This morning, I installed Sourcery G++ Lite on another (Windows XP Pro) computer, and I saw a popup window telling that P&E driver was installed. There was no request to restart the computer. I have not the USB Multilink for Coldfire here, so I can not test if the installed driver is functionnal. Regards, Claude. From carlos at codesourcery.com Mon Dec 11 18:15:27 2006 From: carlos at codesourcery.com (Carlos O'Donell) Date: Mon, 11 Dec 2006 13:15:27 -0500 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <7.0.1.0.2.20061211114322.021c22a0@electro-technica.com> References: <7.0.1.0.2.20061211114322.021c22a0@electro-technica.com> Message-ID: <20061211181527.GL14845@lios> On Mon, Dec 11, 2006 at 11:47:35AM -0500, Claude Sylvain wrote: > - From that point, I compiled a little project and connected to the > target using GDB. > From GDB, when executing a "load" command, there was error(s), and I > have to make many "load" to finally have the command work correctly. > I tryed also to exit GDB, re-execute GDB. Also disconnect/re-connect the > USB Multilink without success. > > - After I have done a Poweroff and Poweron (the following day), all > was working like a charm ! Please read the section "Installing P&E Drivers" of the "Getting Started" guide. It covers the correct sequence to try if you have problems with the P&E drivers. Cheers, Carlos. -- Carlos O'Donell CodeSourcery carlos at codesourcery.com (650) 331-3385 x716 From nathan at codesourcery.com Tue Dec 12 11:19:34 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 11:19:34 +0000 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <7.0.1.0.2.20061211114322.021c22a0@electro-technica.com> References: <7.0.1.0.2.20061211114322.021c22a0@electro-technica.com> Message-ID: <457E9046.8000701@codesourcery.com> Claude Sylvain wrote: Thanks for the detailed report. > - From that point, I compiled a little project and connected to the > target using GDB. > From GDB, when executing a "load" command, there was error(s), and I > have to make many "load" to finally have the command work correctly. > I tryed also to exit GDB, re-execute GDB. Also disconnect/re-connect the > USB Multilink without success. What errors did you get at this point? > Remark: > > - This morning, I installed Sourcery G++ Lite on another (Windows XP Pro) > computer, and I saw a popup window telling that P&E driver was installed. > There was no request to restart the computer. > I have not the USB Multilink for Coldfire here, so I can not test > if the installed driver is functionnal. If you can see if that works before rebooting, that would be helpful. It might be that the P&E install should be requesting a reboot. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From nathan at codesourcery.com Tue Dec 12 11:21:32 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 11:21:32 +0000 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 In-Reply-To: <7.0.1.0.2.20061211115629.02195db0@electro-technica.com> References: <7.0.1.0.2.20061211115629.02195db0@electro-technica.com> Message-ID: <457E90BC.6030006@codesourcery.com> Claude Sylvain wrote: > > - I have the same problem too. > > - Is there a way to get rid of that problem, or do I have to use the > previous version of Sourcery G++ Lite ? We have analysed this problem and determined that there is no workaround. Unfortunately the GCC testsuite had no testcase to cover the circumstances of this bug, so we did not detect it in testing. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From nathan at codesourcery.com Tue Dec 12 15:51:26 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 15:51:26 +0000 Subject: [coldfire-gnu-discuss] Sourcery G++ 4.1-30 for ColdFire now available In-Reply-To: <457D2F09.4070003@codesourcery.com> References: <457A9C70.9090609@blueyonder.co.uk> <457D2F09.4070003@codesourcery.com> Message-ID: <457ECFFE.2060900@codesourcery.com> Steve, > The error message indicates that the PE& library is not installed, or > inaccessible. To track this down, use strace > strace m68k-elf-sprite -i > you should see an open call for libUnit_cfz.so. This is installed in > /usr/lib > > open("/usr/lib/libUnit_cfz.so", O_RDONLY) = 3 have you resolved this issue? nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From csylvain at electro-technica.com Tue Dec 12 16:41:27 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Tue, 12 Dec 2006 11:41:27 -0500 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. Message-ID: <7.0.1.0.2.20061212114010.022177e0@electro-technica.com> > > Nathan Sidwell wrote: > >> >> Claude Sylvain wrote: >> >> - From that point, I compiled a little project and connected to the >> target using GDB. >> From GDB, when executing a "load" command, there was error(s), and I >> have to make many "load" to finally have the command work correctly. >> I tryed also to exit GDB, re-execute GDB. Also disconnect/re-connect the >> USB Multilink without success. >> > > What errors did you get at this point? > - Sorry, the displayed error message have a lot of lines, and I can not remember them. >> >> Remark: >> - This morning, I installed Sourcery G++ Lite on another (Windows XP Pro) >> computer, and I saw a popup window telling that P&E driver was installed. >> There was no request to restart the computer. >> I have not the USB Multilink for Coldfire here, so I can not test >> if the installed driver is functionnal. >> > > If you can see if that works before rebooting, that would be helpful. > > It might be that the P&E install should be requesting a reboot. > - The GDB load command was not working correctly until I have rebooted the computer. - After installing the P&E drivers, connecting the USB Multilink, and followed the Windows add new hardware wizard, there was never a request to restart de computer. - Maybe all this problem is due to the guy that was in front of the computer (that's me). I explain: When I installed Code Sourcery G++ Lite, I did not notice that P&E driver was installed, and I installed another set of driver from the P&E disk that come with the M52235EVB box. It was not a good way to go. Regards, Claude. From nathan at codesourcery.com Tue Dec 12 16:54:57 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 16:54:57 +0000 Subject: [coldfire-gnu-discuss] Connecting to a Coldfire target using P&E USB Coldfire Multilink. In-Reply-To: <7.0.1.0.2.20061212114010.022177e0@electro-technica.com> References: <7.0.1.0.2.20061212114010.022177e0@electro-technica.com> Message-ID: <457EDEE1.9060602@codesourcery.com> Claude Sylvain wrote: > - The GDB load command was not working correctly until I have rebooted > the computer. Thanks. we're checking with P&E as to whether a reboot should be requested during the install process. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From csylvain at electro-technica.com Tue Dec 12 17:19:51 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Tue, 12 Dec 2006 12:19:51 -0500 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 Message-ID: <7.0.1.0.2.20061212121840.02240078@electro-technica.com> > > Nathan Sidwell wrote: > >> >> Claude Sylvain wrote: >> >> >> - I have the same problem too. >> - Is there a way to get rid of that problem, or do I have to use the >> previous version of Sourcery G++ Lite ? >> > > We have analysed this problem and determined that there is no workaround. > Unfortunately the GCC testsuite had no testcase to cover the circumstances > of this bug, so we did not detect it in testing. > - The new version of CodeSourcery G++ Lite support the USB Multilink, but I don't think the previous version support it. I need to connect to the target using the USB Multilink. Is it possible to move the USB Multilink functionnalities to the previous version of Code Sourcery G++ Lite ? Claude. From nathan at codesourcery.com Tue Dec 12 17:27:09 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 17:27:09 +0000 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 In-Reply-To: <7.0.1.0.2.20061212121840.02240078@electro-technica.com> References: <7.0.1.0.2.20061212121840.02240078@electro-technica.com> Message-ID: <457EE66D.20306@codesourcery.com> Claude Sylvain wrote: > - The new version of CodeSourcery G++ Lite support the USB Multilink, but > I don't think the previous version support it. > I need to connect to the target using the USB Multilink. > > Is it possible to move the USB Multilink functionnalities to the > previous version of Code Sourcery G++ Lite ? The earlier version will support the USB multilink, it just didn't include the P&E libraries or drivers. If you uninstall the new version, and reinstall the P&E drivers and the old version, you should be all set. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From csylvain at electro-technica.com Tue Dec 12 18:22:59 2006 From: csylvain at electro-technica.com (Claude Sylvain) Date: Tue, 12 Dec 2006 13:22:59 -0500 Subject: [coldfire-gnu-discuss] CodeSourcery 4.1-30 Message-ID: <7.0.1.0.2.20061212132216.02246cb8@electro-technica.com> > > Nathan Sidwell wrote: > > The earlier version will support the USB multilink, it just didn't > include the P&E libraries or drivers. If you uninstall the > new version, and reinstall the P&E drivers and the old version, you > should be all set. > Great! Regards, Claude. From nathan at codesourcery.com Tue Dec 12 18:34:30 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 18:34:30 +0000 Subject: interrupt routine bug Message-ID: <457EF636.1030805@codesourcery.com> Because of the internal compiler error caused by interrupt routines, we will be producing an update next week. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From david at westcontrol.com Tue Dec 12 19:06:52 2006 From: david at westcontrol.com (David Brown) Date: Tue, 12 Dec 2006 20:06:52 +0100 Subject: Debugging under 64-bit windows Message-ID: <457EFDCC.5050301@westcontrol.com> Is there any sort of support for debugging under 64-bit windows? The P&E drivers won't install, so I can't get anywhere with either my USB debugger or my parallel port type. If anyone else is considering 64-bit windows (or even more masochistic, vista) on a PC used for development, I'd recommend against it. It's a real PITA getting drivers working. mvh., David Brown Norway. From carlos at codesourcery.com Tue Dec 12 19:42:54 2006 From: carlos at codesourcery.com (Carlos O'Donell) Date: Tue, 12 Dec 2006 14:42:54 -0500 Subject: [coldfire-gnu-discuss] Debugging under 64-bit windows In-Reply-To: <457EFDCC.5050301@westcontrol.com> References: <457EFDCC.5050301@westcontrol.com> Message-ID: <20061212194254.GQ14845@lios> On Tue, Dec 12, 2006 at 08:06:52PM +0100, David Brown wrote: > Is there any sort of support for debugging under 64-bit windows? The > P&E drivers won't install, so I can't get anywhere with either my USB > debugger or my parallel port type. > > If anyone else is considering 64-bit windows (or even more masochistic, > vista) on a PC used for development, I'd recommend against it. It's a > real PITA getting drivers working. > > mvh., > > David Brown > Norway. David, Thanks for commenting on your experience with this particular issue. We recommend you use 32-bit Windows for Coldfire development. Cheers, Carlos. -- Carlos O'Donell CodeSourcery carlos at codesourcery.com (650) 331-3385 x716 From nathan at codesourcery.com Tue Dec 12 19:43:19 2006 From: nathan at codesourcery.com (Nathan Sidwell) Date: Tue, 12 Dec 2006 19:43:19 +0000 Subject: [coldfire-gnu-discuss] Debugging under 64-bit windows In-Reply-To: <457EFDCC.5050301@westcontrol.com> References: <457EFDCC.5050301@westcontrol.com> Message-ID: <457F0657.2040505@codesourcery.com> David Brown wrote: > Is there any sort of support for debugging under 64-bit windows? The > P&E drivers won't install, so I can't get anywhere with either my USB > debugger or my parallel port type. I will ask P&E if they have 64 bit drivers. nathan -- Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery nathan at codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk From mark at codesourcery.com Tue Dec 19 18:11:10 2006 From: mark at codesourcery.com (Mark Mitchell) Date: Tue, 19 Dec 2006 10:11:10 -0800 Subject: Sourcery G++ 4.1-32 for ColdFire now available Message-ID: <45882B3E.1050209@codesourcery.com> CodeSourcery is pleased to announce a new version of Sourcery G++ for ColdFire. This release is an update to the recent 4.1-30 release, with fixes for the following critical issues: * An internal error from the compiler on functions using __attribute__((interrupt)) has been eliminated. * A code-generation problem that affected functions using only floating-point registers has been corrected. * A defect in GLIBC's copying of wide characters has been corrected. * Additional CCS examples are included. A confusing error message relating to use of P&E drivers has been improved. Enjoy! -- Mark Mitchell CodeSourcery mark at codesourcery.com (650) 331-3385 x713