Re: [Lazarus] Android LCL example with 2.7.1
Am 07.03.2013 07:25, schrieb Mattias Gaertner: On Thu, 7 Mar 2013 07:07:07 +0100 Sven Barth wrote: Am 06.03.2013 23:53 schrieb "Mattias Gaertner" : On Wed, 06 Mar 2013 15:48:43 +0100 Sven Barth wrote: Hello together! I've managed to run the Android LCL example compiled with the arm-android port of 2.7.1 on the Android emulator (API version 8) with default settings (no -Cparmv6). What is important though is that in the lcl.lpk the following code in the Conditionals is removed (line 49) otherwise a SIGILL error happens when starting the app: === code begin === if TargetCPU='arm' then CustomOptions := '-CpARMV6'; === code end === It can only be removed for fpc 2.7.1? What about 2.6.2? The only compiler that supports arm-android is 2.7.1. Thanks. I removed it. Thank you. Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
Am 06.03.2013 23:28, schrieb leledumbo: Nice news Sven, so now I can drop the hacked 2.5.1? Exactly. Just follow the instructions at http://wiki.freepascal.org/Android to generate a 2.7.1 compiler for arm-android (of course with a bit of /dev/brain0 applied ;) ). Also possible is i386-android. mips-android is AFAIK not supported yet. Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] What's "External: SIGFPE"?
> Could anyone explain what's SIGFPE? SIGFPE stands for SIGnal Floating Point Exception, that is caused by a probably invalid operation involving floating points. > Why there is a SetExceptionMask? Probably to override the same thing already done by (implicitly used) libraries, e.g. gtk2 > Why a normal empty form application will generate SIGFPE with these masks, > and how to trace down to the cause of this error? Because setting the mask means telling the processor to generate (or not to generate? I forgot) SIGFPE for the given operations. e.g. exZeroDivide may trigger SIGFPE if there's a division by zero with floating point operation. How to trace? You already get it, the program will report the line that causes it, you can directly go to the respective line to see what happens. But this requires that the code lies in the program space (not in external library) and the code is compiled with runtime error backtrace (-gl). Otherwise, the program can only report as deep as it can find. -- View this message in context: http://free-pascal-lazarus.989080.n3.nabble.com/Lazarus-What-s-External-SIGFPE-tp4029706p4029715.html Sent from the Free Pascal - Lazarus mailing list archive at Nabble.com. -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] What's "External: SIGFPE"?
2013/3/7 leledumbo > Because setting the mask means telling the processor to generate (or not to > generate? I forgot) SIGFPE for the given operations. e.g. exZeroDivide may > trigger SIGFPE if there's a division by zero with floating point operation. > How to trace? You already get it, the program will report the line that > causes it, you can directly go to the respective line to see what happens. > But this requires that the code lies in the program space (not in external > library) and the code is compiled with runtime error backtrace (-gl). > Otherwise, the program can only report as deep as it can find. > I don't understand: 1) even if I set the masks, why an empty LCL application generate SIGFPE? It means that somewhere GTK or whatever library *indeed* do things like 1/0, only that they are normally hidden/not reported? 2) I already found the problem in my code which is a like look like: *if tbLog.Down then Y := exp(Y);* And Y is a big number e.g. 2500, which caused floating point overflow. The problem was that the number itself should be Log()ed, so that exp() only restore its original value. That was my bug, but the weird thing is that SIGFPE is generated in unit LCLType on the line that set font height, which is very innocent, and after I commented out the font.Height operation, the SIGFPE slipped to somewhere else, but NEVER on the line that generated it! Could you explain why? and how to locate where the SIGFPE occurred? Thanks! -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] What's "External: SIGFPE"?
Most likely it is your GTK+ theme. Many GTK+ themes and theme engines abuse floating point calculations which go unnoticed normally since the vast majority of C/C++ programs do not check for them. The solution is to do something like (IIRC): SetExceptionMask([]); Which disables all floating point exceptions. Alternatively switch to a theme that doesn't mess FP calculations, bu you can't guarantee that your users will have it. On Thu, Mar 7, 2013 at 10:04 AM, Xiangrong Fang wrote: > 2013/3/7 leledumbo > > >> Because setting the mask means telling the processor to generate (or not >> to >> generate? I forgot) SIGFPE for the given operations. e.g. exZeroDivide may >> trigger SIGFPE if there's a division by zero with floating point >> operation. >> How to trace? You already get it, the program will report the line that >> causes it, you can directly go to the respective line to see what happens. >> But this requires that the code lies in the program space (not in external >> library) and the code is compiled with runtime error backtrace (-gl). >> Otherwise, the program can only report as deep as it can find. >> > > I don't understand: > > 1) even if I set the masks, why an empty LCL application generate SIGFPE? > It means that somewhere GTK or whatever library *indeed* do things like > 1/0, only that they are normally hidden/not reported? > > 2) I already found the problem in my code which is a like look like: > > *if tbLog.Down then Y := exp(Y);* > > And Y is a big number e.g. 2500, which caused floating point overflow. > The problem was that the number itself should be Log()ed, so that exp() > only restore its original value. That was my bug, but the weird thing is > that SIGFPE is generated in unit LCLType on the line that set font height, > which is very innocent, and after I commented out the font.Height > operation, the SIGFPE slipped to somewhere else, but NEVER on the line that > generated it! > > Could you explain why? and how to locate where the SIGFPE occurred? > > Thanks! > > > > -- > ___ > Lazarus mailing list > Lazarus@lists.lazarus.freepascal.org > http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus > > -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
[Lazarus] Free-pascal question
Hi all This might be slightly off-topic... I have some fpc -related questions: 1. How good is Lazarus/fpc for embedded programming (compared to C)? It will probably be on a (custom-made) device with ARM/Atmel-based chipset (with or without an OS - will this make a difference?). 2. I remember that the original Turbo Pascal (and Delphi) could use in-line (i386) assembler. Is this still possible with Lazarus/free-pascal and for other chipsets like Atmel/ARM? Thanks Chavoux -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Am 07.03.2013 10:45, schrieb Chavoux Luyt:
1. How good is Lazarus/fpc for embedded programming (compared to C)?
It will probably be on a (custom-made) device with ARM/Atmel-based
chipset (with or without an OS - will this make a difference?).
I can't tell you how good or bad it is, but there are quite some boards
supported already by FPC. You can take a look here for the current list:
http://svn.freepascal.org/cgi-bin/viewvc.cgi/trunk/rtl/embedded/arm/
2. I remember that the original Turbo Pascal (and Delphi) could use
in-line (i386) assembler. Is this still possible with
Lazarus/free-pascal and for other chipsets like Atmel/ARM?
FPC supports inline assembly for each supported target except the JVM
ones. Of course you need to write in the correct assembly language for
each CPU (and maybe differentiate with {$ifdef CPUXYZ} if you want to
support multiple processors).
As a sidenote: For pure Pascal development (and I consider embedded
development part of this) I suggest you to use the fpc-pascal list in
the future.
Regards,
Sven
--
___
Lazarus mailing list
Lazarus@lists.lazarus.freepascal.org
http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 2013-03-07 09:45, Chavoux Luyt wrote: > > This might be slightly off-topic... I have some fpc -related questions: > 1. How good is Lazarus/fpc for embedded programming (compared to C)? FPC works just fine for embedded programming. I haven't tried LCL for such tasks, simply because LCL is to "fat" for my taste. Rather large executables, even for a "hello world" application. Paul Breneman has some excellent information about embedded work with FPC in combination with fpGUI toolkit. A simple 2.5MB download and you have a compiler, fpGUI toolkit, Visual Forms Designer and the experimental Maximus IDE. Due to the usage of fpGUI, executables are nice and small too. I've tested his downloads with a Garmin iQue M5 (WinCE device), and with a Raspberry Pi (ARM Linux). Paul uses different devices. http://www.turbocontrol.com/easyfpgui.htm Regards, - Graeme - -- fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal http://fpgui.sourceforge.net/ -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 03/07/2013 10:45 AM, Chavoux Luyt wrote: 1. How good is Lazarus/fpc for embedded programming (compared to C)? It will probably be on a (custom-made) device with ARM/Atmel-based chipset (with or without an OS - will this make a difference?). What exactly do you mean by embedded ? Possible definitions include: - (hard) real-time - direct hardware access - interrupt programming - code in Flash memory (for "diskless" low-cost hardware) - no OS - no GUI (for "headless" low-cost hardware) - memory space efficiency (for low-cost hardware) - speed efficiency (for low-cost hardware) - cross compiling - remote debugging Some people (especially in the German Lazarus forum) currently are experimenting with cross compiling and remote debugging using a QNAP NAS (headless ARM Linux device using an embedded "diskless" Linux distribution) Others are experimenting with Android (on ARM-Linux) or devices like "Raspberry Pi" (extremely cheap ARM-Linux device constructed for educational use). -Michael -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Am 07.03.2013 11:11, schrieb Michael Schnell: Others are experimenting with Android (on ARM-Linux) or devices like "Raspberry Pi" (extremely cheap ARM-Linux device constructed for educational use). Note: for Android the new native android target in 2.7.1 should be used (currently supports arm-android and i386-android, but AFAIK not yet mips-android). Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 03/07/2013 11:28 AM, Sven Barth wrote: Note: for Android the new native android target in 2.7.1 should be used I understand this is the JVM compiler. Correct ? -Michael -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
[Lazarus] History Window?
Hi, I accidentally pressed Ctrl+Alt+H and it brought up a History window. See attached image. Pressing F1 didn't bring up any help (using Lazarus 1.0.6). What is that window used for? Regards, - Graeme - -- fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal http://fpgui.sourceforge.net/ <>-- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Am 07.03.2013 11:35, schrieb Michael Schnell: On 03/07/2013 11:28 AM, Sven Barth wrote: Note: for Android the new native android target in 2.7.1 should be used I understand this is the JVM compiler. Correct ? No, you're wrong. Since a few weeks native android support was merged to trunk. When I write arm-android, i386-android and mips-android I mean these the native platforms. If I talk about the JVM one that is jvm-android. Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Hi Michael On 7 March 2013 12:11, Michael Schnell wrote: > On 03/07/2013 10:45 AM, Chavoux Luyt wrote: > >> 1. How good is Lazarus/fpc for embedded programming (compared to C)? It >> will probably be on a (custom-made) device with ARM/Atmel-based chipset >> (with or without an OS - will this make a difference?). >> >> What exactly do you mean by embedded ? > I am not going to build the hardware myself, but will probably (help to) program the device. It will be a programmable GPS collar, so I am not sure what the engineers are going to include. But to answer your questions: > > Possible definitions include: > > - (hard) real-time > Probably not necessary. It will need a real-time clock, though. > - direct hardware access > Yes, this will be needed. > - interrupt programming > Yes, probably. I want to include an accelerator meter and it should send a hardware interrupt to tell the device to take a GPS reading. > - code in Flash memory (for "diskless" low-cost hardware) > Definitely diskless to fit in a GPS collar. > - no OS > Although an OS will probably make things easier, low power consumption will be the most (?) important requirement. An OS that runs all the time (or use too much power to boot up every time it receives a hardware interrupt) will probably not work (even if an open-source OS exists for a specific chipset). > - no GUI (for "headless" low-cost hardware) > Yes! (If there is an OS, it might be possible to have a web interface and embedded web server for settings etc., though). > - memory space efficiency (for low-cost hardware) > Yes! It needs to store at least 2160 GPS data points (in addition to the main program itself and any configuration settings). > - speed efficiency (for low-cost hardware) > Yes! > - cross compiling > Yes, the development will be done on Linux (either fpc or c cross-compiler) and written to the device using whatever pins/connectors are available (probably USB and possibly wireless/UHF - this will depend on hardware design) > - remote debugging > Will depend on the chipset and hardware design... can this be done using fpc? .Thanks Chavoux -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 7-3-2013 11:57, Chavoux Luyt wrote: > Will depend on the chipset and hardware design... can this be done using > fpc? fpc list please. I'm sure having this kind of information there would be helpful both to current subscribers and people searching the fpc list later. Having it in the Laz list makes it awkward to search and pollutes the dicussions. Thanks, Reinier -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
Hello, any estimate when the Lazarus daily snapshots will contain fpc 2.7.1. Regards, Sam Von: Sven Barth An: lazarus@lists.lazarus.freepascal.org Gesendet: 9:35 Donnerstag, 7.März 2013 Betreff: Re: [Lazarus] Android LCL example with 2.7.1 Am 07.03.2013 07:25, schrieb Mattias Gaertner: > On Thu, 7 Mar 2013 07:07:07 +0100 > Sven Barth wrote: > >> Am 06.03.2013 23:53 schrieb "Mattias Gaertner" : >>> On Wed, 06 Mar 2013 15:48:43 +0100 >>> Sven Barth wrote: >>> Hello together! I've managed to run the Android LCL example compiled with the arm-android port of 2.7.1 on the Android emulator (API version 8) with default settings (no -Cparmv6). What is important though is that in the lcl.lpk the following code in the Conditionals is removed (line 49) otherwise a SIGILL error happens when starting the app: === code begin === if TargetCPU='arm' then CustomOptions := '-CpARMV6'; === code end === >>> It can only be removed for fpc 2.7.1? >>> What about 2.6.2? >> The only compiler that supports arm-android is 2.7.1. > Thanks. I removed it. Thank you. Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus-- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Hi Reinier On 7 March 2013 13:18, Reinier Olislagers wrote: > > On 7-3-2013 11:57, Chavoux Luyt wrote: > > Will depend on the chipset and hardware design... can this be done using > > fpc? > fpc list please. I am not currently a member of that list. Should I join fpc-Pascal or fpc-devel to ask this? (I assume it is one of the lists found on this page: http://www.freepascal.org/maillist.var)? Regards Chavoux -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
But why is -CpARMV6 a problem? It is correct. The default is ARMv4 which is not supported by Android, so we change it to ARMv6. Looks like a compiler bug or change in the code generator? Also, was the default changed? More details should appear by finding the exact assembler instruction which causes the problem, like I did when I added this directive. -- Felipe Monteiro de Carvalho -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Am 07.03.2013 12:32, schrieb Chavoux Luyt: Hi Reinier On 7 March 2013 13:18, Reinier Olislagers wrote: On 7-3-2013 11:57, Chavoux Luyt wrote: Will depend on the chipset and hardware design... can this be done using fpc? fpc list please. I am not currently a member of that list. Should I join fpc-Pascal or fpc-devel to ask this? (I assume it is one of the lists found on this page: http://www.freepascal.org/maillist.var)? As you want to use FPC (or at least research whether you should) the correct list would be fpc-pascal. And the link is correct. Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
[Lazarus] Error while compiling
Hello, When i try to compile a a simple project, the linker complains: Linking demo /usr/bin/ld: warning: link.res contains output sections; did you forget -T? /usr/local/lib64/libX11.a(xcb_io.o): In function `process_responses': (.text+0x4f9): undefined reference to `xcb_xlib_unlock' /usr/local/lib64/libX11.a(xcb_io.o): In function `process_responses': (.text+0x522): undefined reference to `xcb_xlib_lock' /usr/local/lib64/libX11.a(xcb_io.o): In function `_XReply': (.text+0x9ec): undefined reference to `xcb_xlib_unlock' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XPutXCBBuffer': (.text+0xcb): undefined reference to `xcb_get_request_sent' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XPutXCBBuffer': (.text+0x41f): undefined reference to `xcb_get_request_sent' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XGetXCBBuffer': (.text+0x5f5): undefined reference to `xcb_get_request_sent' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XCBUnlockDisplay': (.text+0x73c): undefined reference to `xcb_xlib_unlock' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XCBUnlockDisplay': (.text+0x760): undefined reference to `xcb_get_request_sent' /usr/local/lib64/libX11.a(xcb_lock.o): In function `_XCBLockDisplay': (.text+0x80c): undefined reference to `xcb_xlib_lock' demo.lpr(20,1) Error: Error while linking demo.lpr(20,1) Fatal: There were 1 errors compiling module, stopping TMessagesView.CollectLineParts WARNING: 28<>8 SrcLine=demo.lpr(20,1) Fatal: There were 1 errors compiling module, stopping I'm on openSuse 12.2, using lazarus-1.0.6-0 and fpc-2.6.0-2. What library am i missing? Thanks -- Ivo Carrajola -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
Think it's a problem because the android emu up to android 2.3 (api level 9) uses ARMv5, not ARMv6. On Thu, Mar 7, 2013 at 12:33 PM, Felipe Monteiro de Carvalho < felipemonteiro.carva...@gmail.com> wrote: > But why is -CpARMV6 a problem? It is correct. The default is ARMv4 > which is not supported by Android, so we change it to ARMv6. Looks > like a compiler bug or change in the code generator? Also, was the > default changed? > > More details should appear by finding the exact assembler instruction > which causes the problem, like I did when I added this directive. > > -- > Felipe Monteiro de Carvalho > > -- > ___ > Lazarus mailing list > Lazarus@lists.lazarus.freepascal.org > http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus > -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
On Thu, Mar 7, 2013 at 9:05 AM, Dejan Boras wrote: > Think it's a problem because the android emu up to android 2.3 (api level 9) > uses ARMv5, not ARMv6. ARMv5 and v6 are nearly identical and I´ve never seen a real device that uses ARMv5, all low end use v6. Also I tested the previous setup in the emulator and also 30+ devices. I doubt that the new setup wont crash in a lot of real devices. -- Felipe Monteiro de Carvalho -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] History Window?
On 07/03/2013 10:45, Graeme Geldenhuys wrote: Hi, I accidentally pressed Ctrl+Alt+H and it brought up a History window. See attached image. Pressing F1 didn't bring up any help (using Lazarus 1.0.6). What is that window used for? F1 should bring you to the wiki http://wiki.lazarus.freepascal.org/IDE_Window:_Debug_History The debugger keeps track of the last 25 times the app stopped. (e.g. breakpoint, finished step). This allows you to view stack, watches etc, as they were then. However, only if you stopped long enough to allow evaluation. If you immediately step again, then not all watches are evaluated. There is also a breakpoint, that will not stop, but take a snapshot. So if you have a focus sensitive app, you can collect all data, and later review it. Finally, it has an export, so you can send your debug session to a fellow developer. (Though there is an issue with clicking stack entries, as pathes are not yet adjusted) -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 03/07/2013 11:57 AM, Chavoux Luyt wrote: Although an OS will probably make things easier, low power consumption will be the most (?) important requirement. If power consumption is a very major issue, the best possible arch that somewhere is in reach is the PIC32 which in fact is a MIPS variant. I know that a MIPS implementation for the fpc compiler is being worked on. - no GUI (for "headless" low-cost hardware) Yes! (If there is an OS, it might be possible to have a web interface and embedded web server for settings etc., though). While no-GUI projects are not too hard a problem, you need to consider, that the normal Delphi like "event driven" programming paradigm is not possible with the current Lazarus library (it is possible with the "mse" library, that is similar on other accounts.) - cross compiling Yes, the development will be done on Linux (either fpc or c cross-compiler) and written to the device using whatever pins/connectors are available (probably USB and possibly wireless/UHF - this will depend on hardware design) While fpc officially does support Cross compiling, it is no especially easy to get it working (see the appropriate Wiki articles) - remote debugging Will depend on the chipset and hardware design... can this be done using fpc? Lazarus uses gdb for debugging. If you don't use an OS (such as Linux) on the target, you can't use the GNU created gdb "receiver" (either gdb or gdbserver). I seem to remember that in this forum I read something about a standard protocol that allows to attach hardware debugger devices, that someone might be willing to implement into Lazarus. -Michael -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Android LCL example with 2.7.1
Am 07.03.2013 12:33, schrieb Felipe Monteiro de Carvalho: But why is -CpARMV6 a problem? It is correct. The default is ARMv4 which is not supported by Android, so we change it to ARMv6. Looks like a compiler bug or change in the code generator? Also, was the default changed? The default for Android was changed to armv5t (though I don't know why Yury chose the *t variant...) More details should appear by finding the exact assembler instruction which causes the problem, like I did when I added this directive. The offending instruction was a "uxth" in customdrawn_androidproc.FPColorToAndroidColor which is only generated by the compiler if the CPU type is ARMv6 or higher (and according to the reference manual for ARM it does indeed only exist for ARMv6 and higher). And this instruction crashed the emulator (which is ARMv5 at least for API version 8). I've also tested the LCL test app compiled for ARMv5 on my Evo 3D which has an ARMv7 processor and the app runs without any crashes. As 2.7.1 officially suppports (arm-)android you can now report bugs if an app crashes because of an illegal instruction, but please state the CPU's model as detailed as possible (for my phone it would be "Qualcomm Snapdragon S3 MSM8260") Regards, Sven -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
On 2013-03-07 12:33, Michael Schnell wrote: > While fpc officially does support Cross compiling, it is no especially > easy to get it working (see the appropriate Wiki articles) All it takes is a bit of practice with the installation. Alternatively, use CodeTyphon [http://www.pilotlogic.com/sitejoom/index.php/codetyphon]. CodeTyphon is a repackaged FPC & Lazarus with lots of cross-compiling options all built-in, plus every component you could ever wish for. It is like the "Studio version" of Lazarus. :) Quite impressive actually. > Lazarus uses gdb for debugging. If you don't use an OS (such as Linux) > on the target, you can't use the GNU created gdb "receiver" (either gdb > or gdbserver). You could also debug using a debugserver and the dbugintf unit included with FPC. Making it communicate over TCP shouldn't be too hard, and is on my todo list. FPC includes a rudimentary console debug server, and LCL and fpGUI versions also exist, which are much better. Regards, - Graeme - -- fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal http://fpgui.sourceforge.net/ -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] History Window?
On 2013-03-07 12:19, Martin wrote: > > The debugger keeps track of the last 25 times the app stopped. (e.g. > breakpoint, finished step). Nice. Thanks for the explanation. > There is also a breakpoint, that will not stop, but take a snapshot. So > if you have a focus sensitive app, you can collect all data, and later > review it. Very handy. I had to always resort to using the "debug server" (and dbugintf unit) to accomplish this. Regards, - Graeme - -- fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal http://fpgui.sourceforge.net/ -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
[Lazarus] [FEATURE] Improves the project inspector
Hello, Please see: http://bugs.freepascal.org/view.php?id=24010 Thank you! -- Silvio Clécio My public projects - github.com/silvioprog -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Error while compiling
Installing libx11-dev should automatically install libxcb-dev (which is the one you miss) -- View this message in context: http://free-pascal-lazarus.989080.n3.nabble.com/Lazarus-Error-while-compiling-tp4029732p4029741.html Sent from the Free Pascal - Lazarus mailing list archive at Nabble.com. -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Ubuntu 12.10 install problem
Hi, I am running lazarus on amd64bit kubuntu 12.10 with kde 4.9.5 This is what I did to get the latest versions. I downloaded the following rpm versions from the lazarus and freepascal sites. fpc_2.6.0-2.fc17_amd64.rpm fpc-doc_2.6.0-2.fc17_amd64.rpm fpc-src_2.6.0-2.fc17_amd64.rpm lazarus_1.0.4-0_amd64.rpm I then converted to deb sudo alien -k fpc-2.6.0-2.fc17.x86_64.rpm sudo alien -k fpc-doc-2.6.0-2.fc17.x86_64.rpm sudo alien -k fpc-src-2.6.0-2.fc17.x86_64.rpm sudo alien -k lazarus-1.0.4-0.x86_64.rpm then installed the deb versions sudo dpkg -i fpc_2.6.0-2.fc17_amd64.deb sudo dpkg -i fpc-doc_2.6.0-2.fc17_amd64.deb sudo dpkg -i fpc-src_2.6.0-2.fc17_amd64.deb sudo dpkg -i lazarus_1.0.4-0_amd64.deb This all went without a hitch and is fully working. This remind me I must update to lazarus 1.0.6. Hope this helps Paul -- ---This message has been sent using Thunderbird on kubuntu--- -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Free-pascal question
Hi On 7 March 2013 15:20, Graeme Geldenhuys wrote: > On 2013-03-07 12:33, Michael Schnell wrote: >> While fpc officially does support Cross compiling, it is no especially >> easy to get it working (see the appropriate Wiki articles) > > > All it takes is a bit of practice with the installation. > > Alternatively, use CodeTyphon > [http://www.pilotlogic.com/sitejoom/index.php/codetyphon]. CodeTyphon is > a repackaged FPC & Lazarus with lots of cross-compiling options all > built-in, plus every component you could ever wish for. It is like the > "Studio version" of Lazarus. :) Quite impressive actually. Thanks Graeme, I'll have a look. >> Lazarus uses gdb for debugging. If you don't use an OS (such as Linux) >> on the target, you can't use the GNU created gdb "receiver" (either gdb >> or gdbserver). > > You could also debug using a debugserver and the dbugintf unit included > with FPC. Making it communicate over TCP shouldn't be too hard, and is > on my todo list. FPC includes a rudimentary console debug server, and > LCL and fpGUI versions also exist, which are much better. Or use a well-placed oscilloscope. ;-) I have done a few embedded projects in the past (using C/assembler) where that was my only debugging tool... not so much fun for a programmer, but a good learning experience. Unfortunately, I don't have my own oscilloscope (they are really expensive), but I hope one of the engineers will have one. Cheers Chavoux -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
Re: [Lazarus] Ok I give up!
On 3/5/2013 13:01, Jürgen Hestermann wrote: I just recently installed a snapshot version of Lazarus on Windows (I cannot copy the version number from the "about" window because I cannot select text. Why?). right click on the help window where the graphic is ;) -- ___ Lazarus mailing list Lazarus@lists.lazarus.freepascal.org http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
[Lazarus] SQLite - disk I/O error only on virtual machine
Hi,
I noticed strange behavior when running application on virtualbox linux
mint 14 64bit machine. This occur only when I attach second database to
sqlite3connection and then call SQLTransaction.Commit. This is my code:
try
FConn := TMySqlite3Connection.Create(Self);
FConn.Transaction := SQLTransaction1;
FConn.DatabaseName :=
IncludeTrailingBackslash(ProgramDirectory)+'db1.sqlite';
FConn.Connected := True;
FConn.ExecSQLNoTrans('attach database '+
QuotedStr(IncludeTrailingBackslash(ProgramDirectory)+'db2.sqlite')+'
as d2');
FConn.ExecuteDirect('create table if not exists a(col integer)');
FConn.ExecuteDirect('create table if not exists d2.b(col integer)');
SQLTransaction1.Commit;
except on e: Exception do
ShowMessage(e.Message);
end;
Even if error occur, commit succesful created tables. When I run it on
normal Linux installation (also Mint 14 64bit) then everything is ok. Both
OS have same sqlite version (3.7.13) from ubuntu repo.
I will post this report on sqlite mailing list too, but maybe you had
similar problem.
Regards
--
___
Lazarus mailing list
Lazarus@lists.lazarus.freepascal.org
http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
