Re: [fpc-pascal] Re: Converting code from C++ to FP....
Jorge Aldo G. de F. Junior wrote: const has the same effect as fpc const before a parameter, This is not true. double* SX0 is the same as double *SX0 or double * SX0 so this becomes const SX0 : pdouble (ugly pointer as array Cishism) No! 'const double * X' in C makes the pointed-to data const, while 'const X: PDouble' in pascal makes the parameter itself const. You cannot assign another address of a double to it, while you can in C. In C terms, pascal can only do 'double * const X'. Pascal is not really useful this way, 'const' is generally only useful for structures and strings. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] GetTempFileName in Linux
Michael Van Canneyt wrote: I haven't seen any yet. The main problem is that the update isn't 'optional' or 'to be scheduled at X every night'. It must be done when the server application says it is time to do so. apt-proxy ... push to the proxy when it's time to update stuff ;) Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] GetTempFileName in Linux
Michael Van Canneyt wrote: Yes, but if the administrator must update 20 PCs manually every 2 weeks, he quickly complains that he has better things to do, so it must be done when the 'ordinary' user uses the application. Aren't there automation systems for this? Just like debian's 'apt-get update && upgrade' with a custom repository with your software? But then for windows I presume? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FPC static linking of zlib
Michael Van Canneyt wrote: libz.so.1 => /lib/libz.so.1 (0x7f6de2801000) As you can see, it uses libz (or zlib) but dynamically, probably through some library dependency. I think this dynamic library dependency is caused by e.g. gtk, not directly by application or lcl or fcl directly. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FPC static linking of zlib
Matthias Klumpp wrote: If you build the package using lazbuild, lintian (the Debian policy checker tool) throws an error described here: http://lintian.debian.org/tags/embedded-zlib.html paszlib is a pascal implementation of compression. How is that check performed? Maybe it triggers on some specific function name? I'm not sure if some version of zlib was translated to pascal; therefore having the same security issues, or if it was written from scratch, so that it won't have those security issues? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Pointers
Rainer Stratmann wrote:
Ok, that makes some sense, but I did not know it before.
In the past with the turbopascal compiler and other always sizeof byte was
added.
The behavior is dependent on the {$T+} (typed pointers) mode.
Micha
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] readonly variables
Anthony Walter wrote: Having said all that, Jonas, what is the actual implemented behaviour of FPC? Does it 0 initialize heap memory at startup or not? If not, what is the justification for not doing so when this has been a long established behaviour of Delphi? It's not the compiler or RTL that zeroes the memory, but the OS. Therefore, it cannot be guaranteed by the fpc documentation (except if it would contain code to explicitly initialise that memory). Most OSes (Windows, Linux, BSDs, ...) do zero global memory though, to prevent leakage of (potentially security sensitive) information from one process to another. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Generic type declaration example versus manual.
Jonas Maebe wrote: tgeneric16.pp is not an example, it's a test. Moreover, it doesn't compile yet, not even with FPC 2.5.1. So documenting that syntax is not really a good idea. His comment doesn't really apply to tgeneric16 specifically, but more to generic syntax in general. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] bzip2
Juha Manninen wrote: Oops, sorry. I will do that. Where does the thread information come from? It looks good in my email client and I don't see any extra data. Try to enable 'Threaded view'; usually somewhere in the View menu, Sort menu, etc. Depends on client used. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] memory mapped file
Alexey Voytsehovich wrote: All the good days. You can not tell what I need to read and/or study for solving cross-platform (linux/windows) implementation of the memory mapped file? On unix, it would be fpmmap. On windows, it would be CreateFileMapping/MapViewOfFile if I'm not mistaken. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] (no subject)
ak za wrote: i want to don't send mail to me from you but idn't know how do it. please guide me Do you mean unsubscribe, see link at bottom of this email: http://lists.freepascal.org/mailman/listinfo/fpc-pascal Then look at bottom of page, unsubscribe. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Illogical automatic dereferencing
Henry Vermaak wrote: One thing I think you don't understand is that an array _is_ a pointer. Look at this table to visualise: In Pascal, an array is not a pointer; at least not at the language level. For a static array X (array[1..n] of T), you *can* write: Move(Ptr^, X, sizeof(X)); because X *is* the array, representing the memory of the array. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FIONREAD FPC FpIOCtl for Cross platform Sockets
Andrew Brunner wrote: Sockets programming often requires a poll for how much data is available on a particular socket descriptor. AFAIK, it doesn't? If data is available, but less than you request, a 'read' on the socket does not block, but returns the number of bytes read, which is less than you requested. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] multiple inheritence
David Emerson wrote: lists containing the same elements: one sorted, one unsorted. The reason to have both: sorted -> fast search; unsorted -> sequential navigation through the list while elements are being added. It's not as if combining those gives you the best of both worlds ... if the sorted list needs to keep its list sorted, having the sequential list won't make iterating faster ... Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] getting started with threads
Jonas Maebe wrote: on x86 too). Atomic operations are not memory barriers by themselves, and the fact that you perform an atomic operation does not mean that afterwards all cpu's will immediately see this new value. Explain? Isn't the point of an atomic update that it doesn't matter whether the other CPU has seen the value yet? Or are we talking about different values? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: getting started with threads
Graeme Geldenhuys wrote: So far I have single interface for create/destroy/lock/unlock of semaphores for Windows, Linux, *BSD. The latter two is actually for all A semaphore is not "locked" and "unlocked", it is posted and waited for. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus settings and roaming profiles in Windows
Florian Klaempfl wrote: Micha Nelissen schrieb: Florian Klaempfl wrote: And applications can implement a configuration file lock much easier if they want to detect multiple activity. Do they :)? Does lazarus? IIRC OpenOffice (3.0) does detect it already and warn me. Well, then you need two 1k Euro servers. Anyways, a server not responding quickly enough to serve a profile will be a real pain serving a fully networked home/profile dir :) That's the difference between a Windows server and one with a proper OS :-P. Or it's the administration of it. Or ... don't know. I know there are little problems with the file server I described in another post but many problems with corrupted profiles. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus settings and roaming profiles in Windows
Vincent Snijders wrote: I am not sure having a 100 MB lazarus *roaming* profile by default would be a good idea. Probably the choice to compile lazarus itself to its "configuration" directory is not such a good idea ;-). Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus settings and roaming profiles in Windows
Florian Klaempfl wrote: My conclusion: I agree with Jonas, you can use network shares for development machines.. Consider 10-50 people working and accessing a server this way :) Using roaming profiles and local source checkouts, this can be handled easily by a 1k Eur server which even doesn't requires high We have like 150+ people accessing all project documents, designs, code on a single file server. Server runs Linux with Samba AFAIK. Some of Linux targeted development checkouts are on another Linux development server accessed by a small group of people though, but it still proves the occasional "home drive" configuration file on a networked drive could be handled easily. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus settings and roaming profiles in Windows
Florian Klaempfl wrote: Then you must be the only one since everybody I know despises it. We use multiple times login sometimes and then the settings are always messed up. Well, this causes a headache with fully networked home drives as well ;) If configuration files are per application it's much less of a problem. And applications can implement a configuration file lock much easier if they want to detect multiple activity. It reduces the per-login problem to a per-application-start problem; which is much better already. If the profile locally exists already it sometimes isn't updated but afterwards it does overwrite the server's! Hmm, I know only of this if the a machine starts up with a messed up clock because e.g. of an empty mainboard battery. Or a server that is not responding quickly enough or so ... Perhaps a 1k euro server isn't sufficient after all ;-P Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus settings and roaming profiles in Windows
Florian Klaempfl wrote: Roaming profiles using auto copy/delete are not the best choice for that. For me it works fine for years ;) Then you must be the only one since everybody I know despises it. We use multiple times login sometimes and then the settings are always messed up. If the profile locally exists already it sometimes isn't updated but afterwards it does overwrite the server's! The only "solution" is not logging off, wow great engineering indeed! Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] DirectFB for FPC beta tarball
Michael Van Canneyt wrote: The known major bugs of previous SVN versions have beeen fixed. We're now starting to actually use DFB in our project. Reports (bugs or anything) are thus highly appreciated. Do you think it would be feasible to use this as a basis for a Lazarus widgetset ? Lazarus without X would be quite cool. I already have it running, using gtk2 :P Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Class procedure assigned to object's event
Antonio Sanguigni wrote: For the second one: main.pas(261,19) Error: Incompatible types: got "" expected "" Remove the 'class' keyword from your function. A class method is a different beast (with different signature) from a regular method. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] re: thanks for major signal handling fix
Jonas Maebe wrote: On 20 May 2009, at 22:01, Jonas Maebe wrote: The problem you have right now is that the program and each of your libraries each contain their own copy of the exception class, and therefore do not recognise (Pascal) exceptions raised by any of the others. Well, that and the fact that moreover each library/program contains its own exception stack rather than a single shared one (although that could Aren't both problems solved by using the same RTL binary code (in library form)? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] heap size growing
Luca Olivetti wrote: Seeing this thread, I used GetFPCHeapStatus and, effectively, CurrHeapSize is growing but CurrHeapUsed isn't, so apparently TJpegImage.LoadFromStream (that's what I'm using) causes heap fragmentation. Can you reproduce this in a small test program? Perhaps there is a simple fix for TJpegImage's allocation pattern. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Dependencies of FPC_HAS_FEATURE_XXX
Sven Barth wrote: Hmm... if this speeds things up: shall I enclose the problematic places with ifdefs and send a patch along with the bug report? Got response yet? If not then yes, please send patches per topic to ifdef the optional code away. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] REQ: official FPC's FCGI framework
Bee wrote: Hi all, FPC has pretty usable support for web based development. Currently FPC supports it through CGI and Apache module. IMO, there's one more support that should be added officially: FastCGI. The initial work has been done lNet has fastcgi support. Maybe it's interesting for you. It can connect to some server or spawn some itself. Implemented in lib/lfastcgi.pp. It's used by lwebserver to provide a fastcgi connection to PHP backend. In lib/lwebserver.pp. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Using FPC MM in a C program
leledumbo wrote: Since it's possible to use libc's MM via cmem unit, would the opposite be possible? I have a "crtl.pas" which defines a few functions (malloc, calloc, free, memmove, memcpy) to help in using libraries written in C with fpc programs without needing external dependencies (msvcrt.dll on windows, for some reason that did not work for me). However, it's a all-or-nothing situation. A program uses either this crtl unit, or it links a C library like msvcrt.dll or something similar; using both is asking for trouble. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] fstat usage
Francisco Reyes wrote: Trying the fstat function and don't seem to be getting the right values for ctime, mtime and atime. What OS and CPU ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Lazarus: graphical interface using code only?
Graeme Geldenhuys wrote: I'm curious... How do you manage to create a GUI or use a GUI program if you are blind? Sorry, I don't know how screen readers work? Try http://en.wikipedia.org/wiki/Screen_reader for a start? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Make FPC closer to hardware
leledumbo wrote: Would it be a good idea to lower down (again) FPC so that it can be used easier for systems programming? example? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Speed difference between intel/ppc on intel Mac
Jonas Maebe wrote: You can use -vd to have the compiler print out all compiler options it interprets (both from configuration files and from the command line). In general, to get the best code for an Intel Core family processor, use something like -O3ppentium4 -Cppentium4 -Cfsse2 i doubt it'll speed up your program by a factor of 2 in any case though. Hmm there are weak spots in FPC. E.g. fannkuch in the shootout: http://shootout.alioth.debian.org/u32/benchmark.php?test=fannkuch&lang=fpascal&id=5 FPC is 2.1 times slower than gcc here. Maybe this program is like it? It's still quite unlikely that an assembly emulator would optimize it run so much better than the original compiler could. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Translate C to Pascal
Mattias Gärtner wrote:
> How to translate this:
>
> struct a;
Isn't this a forward declaration? So sometime later it needs to declare
'struct a { ... };' ?
Micha
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Enabling FPC_HAS_FEATURE_CONSOLEIO
leledumbo wrote:
> This is very inconvenient. Moreover, assertions needs it. My first attempt,
> is to comment out {$undef FPC_HAS_FEATURE_CONSOLEIO} in my system.pas. But
> then, FPC argues about Unknown compilerproc "fpc_write_text_shortstr". I
> found it in text.inc, but it's only included if FPC_HAS_FEATURE_TEXTIO is
> defined. Knowing that I don't yet have any file system driver, I guess it's
> impossible to reach my goal by defining it.
> What's the correct way?
Can you undefine FPC_HAS_FEATURE_TEXTIO as well?
In the end, you have nothing left :-). Or is this the point?
What's the goal?
Micha
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Generics
Damien Gerard wrote: >> I tried to use an Interface but I had a nice Segementation fault :) >> >> Should I report it as a bug in the mantis ? >> >> > > Actually there is the same error with an intermediate class :( Segmentation faults in the compiler itself can/should always be reported. Thanks. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Generics
Damien Gerard wrote: > Oki I understand. > I tried to use an Interface but I had a nice Segementation fault :) I'm not sure how FParent.ItemByDefault is going to work? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] AnsiDequotedStr not return empty for ""
Peter Vreman wrote: >> In case always check Delphi documentation to see wether the behaviour >> you see is intentional. > > In case of "" it is also ambigious what you want. Because "" in the > middle of the string is translated to a single ". So you might also > expect that it returns a single " instead of an empty string. Wouldn't it be reasonable to expect the rules are similar to Pascal itself, with being a single quote, and '' an empty string ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] AnsiDequotedStr not return empty for ""
Peter Vreman wrote: >> That also in Delphi, but it is for me wrong. >> In fact that make my program working wrong when send UserName="Admin" >> and Password="", it check password as '""' not empty. > > The behaviour is compatible with Delphi and will not be changed. Might as well remove it then, as it seems to me it's pretty useless this way, and people can only get confused in case empty things are not working as expected? Is there a use case in which you really want the empty quoted string to not remove the quotes? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] syscalls and fpc
ik wrote: > Actually the assembler is not that hard to understand :) > My point is, that I don't like the idea of 7 or 20 or 100 amount of > parameters to give answer to every need. I think we should find a > better way to implement it, like var args in C or open array in > pascal... The linux kernel interface is always using registers, and since there are a limited amount of registers, there is a max. number of arguments. IIRC the maximum is 6, i386 limited: eax, ebx, ecx, edx, esi, edi, ebp. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] syscalls and fpc
ik wrote: > Hi, > > I found that the Do_Syscalls are written in assembly and have only > limited number of parameters (up to 6). Is there a way to write it > using array of TSysParam instead of having 7 different functions ? Maybe but it wouldn't make the assembler easier to read :-). > Another question is, is there a way to use it witthout writing > assembly, like using it in libc or some other way ? Calling them doesn't require assembler, or what do you mean? You can try to extend the compiler to support inlining of assembly functions ;-). Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] static class methods
Jonathan Benedicto wrote: > Jonas Maebe wrote: >> I'm not sure how Delphi could handle this without passing the VMT as >> invisible parameter (and from a cursory look at the code generated by >> Kylix, it does appear to pass the VMT). > > Delphi doesn't allow static class methods to be virtual. What are "static" class methods? "Plain" class methods are allowed to be virtual, surely. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Problems with generics
bartek wrote: > I am not quite sure whether my synatx is correct. IMHO explictly defining the > type feels more pascalish. Having more time at hand i have written a much > shorter example code which shows in what way such a type should be defined. I think that it is the same type in a different namespace. Like using "unit.type" syntax. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Problems with generics
bartek wrote: > function HashOf(const Key: ShortString): DWord; > > function BucketNodeOf(const Hash: DWord): TBucket.PNode; > DataContainers.pp(104,66) Fatal: Syntax error, ";" expected but "." found I'd say this is a bug, and your syntax is correct. Update the existing bug report, perhaps? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Heaptrace and generics
bartek wrote: > Could be FPC at fault? Both files are attached, "DataContainers.pp" being the > two lists and "DataContainers_Test.pp" the test program. > > FPC 2.2.0 on Linux 2.6.23 > valgrind-3.2.3 Can you retest with fpc 2.3.1 ? There have been some heaptrace fixes. Otherwise, please post a bug report with the code, so it won't be forgotten. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] CSV via PCRE
Graeme Geldenhuys wrote: > OK, while we are busy with show-and-tell... Then have a look at my > token library implementation. You've implemented some kind of 'cut'. But grep is also very useful (and more often used in a shell, at least by me). Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] fpnanosleep (again)
John Coppens wrote: > Just an extra note: delay(1) seems to delay 4.1 ms, so it _does_ seem to > be possible to do less than 8.3? You wait till the next tick that is after the requested delay. So it's never a HZ multiple or so. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] memory aligning
David Pethes wrote: > pointer. So I mixed the two codes and come with something like this: > > function evk_malloc (size: longword): pointer; > const > ALIGNMENT = 16; > var > ptr: pointer; > begin > ptr := getmem(size + ALIGNMENT); > result := Align (ptr, ALIGNMENT); > if result = ptr then > pbyte(result) += 16; Why not use ALIGNMENT here also? > //store offset to original address > (pbyte(result) - 1)^ := result - ptr; > end; > > Is this correct to use? Wouldn't be there any catches for example when > running on 64-bit machine? This looks ok. BTW, GetMem already returns 16 byte aligned blocks. If it does not, it's considered a bug. > Also, does FPC support aligning stack variables and record members on > mod16 addresses like GCC > (http://gcc.gnu.org/onlinedocs/gcc-4.2.2/gcc/Variable-Attributes.html#Variable-Attributes) Don't know. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: Why this evaluates on "if" wrong ?
Stephen Dickason wrote: > In binary that is 0001 + 0.0110011001100 - 0.0110011001100 because > we hit recurring decimals a lot more in binary than decimal. I wonder why we > don't have a standard format (maybe we do?) that factors in the remainder as > part of the number also? It's possible to create a type that stores the numerator and denominator, but then you would need to "simplify" to extract common factors on every calculation, otherwise it would quickly run out of range (Integer or whatever you choose). Extracting common factors takes exponential time, so it's quite slow. The range would be quite limited as well, it's as wide as the numerator. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FPC and JAVA
Jonas Maebe wrote: > go for any Lazarus/Delphi app ported to the Mac, for that matter (unless > Lazarus can do automatic button layouting, and if the current layout > mismatch is simply due to some wrong default setting for the Carbon > target). Actually, we have a TButtonPanel exactly for this purpose. It's just that (A) not all dialogs use it yet (B) it doesn't get the right default setting from Carbon yet. Perhaps not very easy fix, but the infrastructure exists. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] random "Disk Full" exceptions
Marc Santhoff wrote: >> writeln(stderr, '--> hit found: '+inttostr(fcnt) + ' - ' + >> inttostr(PInteger(longint(adr)-longint(fx))^) + ' - ' + >> floattostr(PSingle(adr)^)); > > Very nasty thing: > > The output channel is temporarily blocked and gives back "EAGAIN". This means the output channel is opened non-blocking. Is that intentional or a bug ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Audio development and /dev/dsp
ik wrote: > lsof does not give me any open access to the device. > > I'm using Kubunto 7.04 amd64. Buggy audio driver ? Does it work properly when using ALSA, aplay etc. ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Problem with interfaces
Joao Morais wrote: > The first call that I make to DefaultObj from *another* unit recreate > the object, because, I don't know why, the _holder var points again to nil. Try to use a data watchpoint to watch when the variable is being changed. Hopefully that will point to the root cause. Did you also try printing the address of the variable (I assume global variables aren't moving, but who knows?) Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] msedb and fcl-db benchmarks
Michael Van Canneyt wrote: >> Maybe you can poll /proc/$PID/status > > Why not simply use heaptrc ? It prints a summary ? Or even GetFPCHeapStatus.MaxHeapUsed. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Linux Kernel and FPC
Michael Van Canneyt wrote: >> And I was thinking in implementing it using FPC. >> My question is, how hard can it be ? > > Relatively hard: you must write a complete RTL for it. > Once that is done, the rest should be easy. I don't think you would want to use a regular pascal RTL in a kernel module. I'd say "just" translate the kernel C headers for interfacing, and arrange the linking process to work well. Compile without system unit. Use mode fpc to prevent usage of objpas unit, although perhaps it doesn't have real dependencies on system unit. Disadvantage of linux kernel is that its API changes a little bit with each release, so keeping the bindings up to date may require more maintenance than other libraries. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] PowerPC crosscompiler on i386 Mac OS X looks for, wrong assembler and linker
Adriaan van Os wrote: > build and then the Makefile takes cares of the rest. If a configuration > is not supported or if there is a problem, you get a clear error message. You *must* be joking, right ? Usually it involves digging through a messy config.log, where you sometimes can't really find the reason (just one of those mysterious test compiles failed) and then have to decrypt an enormously complex configure script that will give you nightmares for years. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Starting new threads
Tom York wrote: > By the way is NNTP forbidden? I would much rather use NNTP over a > mailing list. IIRC, these lists are mirrored in gmane, so perhaps you can use gmane ? Have the mailing list subscription, but configure it to not send email. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: put data to stdout (fpc vs. gcc)
Michael Van Canneyt wrote: > You would have to test the speed of each of them separately to be able > to say anything definitive. Like it is now, there is no indication > that it is the float-to-text routine which is to blame. So use sprintf in C, and the equivalent in fpc; and rerun the test. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] PASCAL programming for Novice
Francisco Reyes wrote: > Pete Cervasio writes: > >> Contents of $HOME/.vim/filetype.vim: > > Thanks. Found that it also works with .pas > Changed the programs to that since that is one of the extensiosn the > compiler also checks for. ___ If you google for "vim delphi syntax" you can easily find a delphi syntax highlight file for vim. You can then connect .pas/.pp/.inc to that one in a similar way. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Linux Linking Error solved
Rick Seiden wrote: > I emailed yesterday about an error when compiling in Linux. The error I > received was "Error while linking." > > I figured out that the permissions on /usr/bin/ld is set to root, and > I'm logged in as a different user. When I sudo fp to start Free Pascal, > I am able to compile without a problem. What are the permissions exactly ? You should be able to execute ld as normal user at least. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] TThread.Suspend under Linux
Graeme Geldenhuys wrote: > I got most of my Thread Manager demo written. While trying out a few > features I noticed that every time as Suspend a thread, my app is > killed. Running it through gdb I get the following output. I don't see it being killed from your output, or am I missing something ? man 7 signal: SIGSTOP 17,19,23StopStop process It stops ... and that's what you asked it to do. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] problem with threads and backgrounding, fpc 2.1.3 arm native
Luca Olivetti wrote: > This is another test program that shows my problem with fork: as soon as > I start the thread, the main program stops working (the thread *is* > running, you can put a writeln in its loop to show it). Note that in > this short example I don't close stdin,stdout,stderr, I don't block > signals but it doesn't matter (I did it all and the result is the same). > The same happens with the native compiler and the cross compiler. I can reproduce this, (gdb) bt #0 0x401afd80 in __pthread_sigsuspend () from /lib/libpthread.so.0 #1 0x401aea7c in __pthread_wait_for_restart_signal () from /lib/libpthread.so.0 #2 0x401afc40 in pthread_create@@GLIBC_2.1 () from /lib/libpthread.so.0 Seems like it's waiting for some signal to be passed to the process, but it doesn't happen. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] problem with threads and backgrounding, fpc 2.1.3 arm native
Luca Olivetti wrote:
>> Backgrounding is not the same as forking.
>
> I (vaguely) understand the difference, but I think (probably mistakenly)
> that the problems are related.
I'm not saying the problems are not related; they may be. It's just
important to know there is a difference.
> the example daemon.pp, for example, crashes right away, strace doesn't
> seem very useful to me:
>
> [EMAIL PROTECTED]:~/fork$ strace -f ./daemon
> execve("./daemon", ["./daemon"], [/* 22 vars */]) = 0
> syscall: unknown syscall trap 0x8000
There is a patch on the internet for strace, to skip this error. Google
for strace syscall trap.
> A simple program that just repeatedly sleeps after the fork, dies as
> soon as it receives any signal (e.h HUP, USR1).
Ok, this will be the same problem as with backgrounding in the shell.
Micha
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] problem with threads and backgrounding, fpc 2.1.3 arm native
Luca Olivetti wrote: > [EMAIL PROTECTED]:~$ uname -a > Linux lspro 2.6.12.6-arm1 #77 Tue Dec 5 22:23:46 CST 2006 armv5tejl > GNU/Linux I've gotten access to an arm-linux system, where I cannot reproduce the crash. That system is using 2.6.18 kernel. Can you try a newer kernel as well ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] problem with threads and backgrounding, fpc 2.1.3 arm native
Luca Olivetti wrote: > synchronize and critical sections, c library with callbacks to pascal, > etc.) is working flawlessly in the foreground, but it segfaults if I try > to background it (or try to use fpFork), and this is a problem since I > want to write a daemon (i can use screen though if that isn't possible). Backgrounding is not the same as forking. Backgrounding is using the shell's job control (and sends some signal(s) to the process), forking is what the program does itself. Can you list an example that crashes with fpfork, and give an strace with it ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
Florian Klaempfl wrote: >> I am finished. It can be improved, but I'm out of ideas =) >> > > I'am not ;) The wiki is definitively not the FPC documentation. I've changed it to 'documentation for developers'. Covers the content better ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] fpc have problem with dynarray in ARM
Jose Pascual wrote: > Could it be resolved in 2.1.1? > > I'm compiling ppcrossarm 2.1.1 in order to be it's corrected in this > version, Open a (descriptive) bug report (in case there is none about this issue yet) and attach a test case. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
Felipe Monteiro de Carvalho wrote: > Hello, > > Are the wikis being merged? Is it my impression, or is there a main > page for compiler things missing? The pages are there. > When I go to fpc page and click on Wiki it shows lazarus-ccr wiki, and > I can´t find what was the old free pascal main wiki page, and thus > can´t find the compiler related articles. Correct, somebody needs to rewrite the Main_Page to mention both projects, content, and links. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Question on how to avoid memory trouble
Wayne Sherman wrote: > Getting a better containers unit is nice, but the root of the problem > seems to be the memory manager. Delphi's old memory manager had No. > fragmentation issues also. That is why NexusMM, BucketMM and FastMM we The memory manager cannot fix fragmentation in TStringList. Delphi's memory manager fragmented memory with (repeated) different memory size usages. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] GetEpochTime not found
Marco van de Voort wrote: >>> fpc 2.0.2. >>> >>> CommonTypes.pp(207,21) Error: Identifier not found "GetEpochTime" >> http://www.freepascal.org/docs-html/rtl/oldlinux/getepochtime.html > > He uses FreeBSD. Ah right, this was apparent from the unit path. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] GetEpochTime not found
Marc Santhoff wrote: > Hi, > > I've got some funny problem here. The program compiled well with fpc up > to version 1.9.4. I haven't touched it for a while and now it fails with > fpc 2.0.2. > > CommonTypes.pp(207,21) Error: Identifier not found "GetEpochTime" http://www.freepascal.org/docs-html/rtl/oldlinux/getepochtime.html Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Implementing a true Singleton - Can we decrease the visibility of a method?
Daniël Mantione wrote: > > Op Fri, 8 Dec 2006, schreef Micha Nelissen: > >> Btw, I think singletons are nonsense too. Why is a global variable evil, >> and a singleton class not ? > > Well, a singleton can hide and/or protect its private data. Not more than a global variable of the same class type ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Implementing a true Singleton - Can we decrease the visibility of a method?
Michael Van Canneyt wrote: >> Try this. >> http://bdn.borland.com/article/22576 NewInstance/FreeInstance is what I would have recommended as well. > Maybe we should include an implementation in FPC by default ? How ? The article works as-is AFAIK. Btw, I think singletons are nonsense too. Why is a global variable evil, and a singleton class not ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] even linus torvalds prefer pascal over c
John Coppens wrote: > Mind, I don't want to remove gotos. I'm all for free choice. But I'm > still very much a proponent to discourage its use for fledgling > programmers. Well, in fpc, you have to use a compiler option to be able to use them. A good trade-off IMHO. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] sparse files
David Mears wrote: > is there a generic multiplatform way to detect sparse files besides > looking for run-ons of #0s? Is this not good enough ? It's at least cross-platform ;-). Also the only one AFAIK, but not sure. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] even linus torvalds prefer pascal over c
Daniël Mantione wrote: > > Op Sat, 2 Dec 2006, schreef Micha Nelissen: > >> Daniël Mantione wrote: >>> (1) I'm using the in English not so common "informatics" instead of >>> "computer science", because Dijkstra greatly disliked the term). >> How about "computing science", or is that too "Eindhovens" ? ;-) > > Dijkstra disliked the term because informaticians/computer scientists > don't study how a computer works. "Informatics is no more about computers Indeed, that's why I said "comput*ing*", i.e. the working with computers, but not the science of the computers themselves. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] even linus torvalds prefer pascal over c
Daniël Mantione wrote: > (1) I'm using the in English not so common "informatics" instead of > "computer science", because Dijkstra greatly disliked the term). How about "computing science", or is that too "Eindhovens" ? ;-) Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] even linus torvalds prefer pascal over c
Daniël Mantione wrote: > Alas, he doesn't. Linus Torvalds stronly disapproves of Pascal and this > statement should be viewed in that context: he wants a language that does > allow him to do anything wrong. Actually: he wants a language to allows him to do things which are perceived by some to be wrong, but are not, after all. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] even linus torvalds prefer pascal over c
Bisma Jayadi wrote: > Linus concluded, "the C language has scoping rules for a reason. *If I > wanted a language that didn't allow me to do anything wrong, I'd be > using Pascal.* As it is, it turns out that things that 'look' wrong on a > local level are often not wrong after all." Well, Linus is not too fond of "standard Pascal" (but neither am I ;-) ), read the following about goto and labels: (http://kerneltrap.org/node/553/2131) On 12 Jan 2003, Robert Love wrote: > On Sun, 2003-01-12 at 15:22, Linus Torvalds wrote: > > No, you've been brainwashed by CS people who thought that Niklaus > > Wirth actually knew what he was talking about. He didn't. He > > doesn't have a frigging clue. > > I thought Edsger Dijkstra coined the "gotos are evil" bit in his > structured programming push? Yeah, he did, but he's dead, and we shouldn't talk ill of the dead. So these days I can only rant about Niklaus Wirth, who took the "structured programming" thing and enforced it in his languages (Pascal and Modula-2), and thus forced his evil on untold generations of poor CS students who had to learn langauges that weren't actually useful for real work. (Yeah, yeah, most _practical_ versions of Pascal ended up having all the stuff necessary to break structure, but as you may be able to tell, I was one of the unwashed masses who had to write in "standard Pascal" in my youth. I'm scarred for life). Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Nov. 30th Build of 2.1.1 for Sparc
Wolfe, Robert wrote: > I would like to port FPC over to FreeBSD for Sparc as well if at all > possible, however, with a lack of an existing FPC compiler for that > platform, I was not sure how I would go about doing that. The compiler is only per architecture, so you only need to build a freebsd rtl using for example a sparc-linux compiler, then build a sparc-freebsd compiler using the freebsd RTL. Also see the buildfaq. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Compiling problems
Jonas Maebe wrote: > >> If it's later fixed people might think, it's still not fixed. > > That should only be changed after it's fixed I think, because if you > write "5091 to 5114" and it's still not fixed in 5115, people may also > be confused :) It just says those revisions are unusable, not whether the others are or are not. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Progress Bar in different widget sets
Graeme Geldenhuys wrote: > GTK's progress bar has a feature where the progress bar moves left and > right repeatedly, instead of the normal from Min to Max. > > What do you call that feature and what do you actually use it for? It's used when the Max isn't known. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Linux Install: Is it me or the documentation? ;-)
Lev Lafayette wrote: > and like the TurboPascal critter from all those years ago, verdade? If Yes. > this is so, where does it live in the filesystem? Not that coding in > vi(m)/emacs/pico etc is a problem as such... /usr/bin/fp ? It's named 'fp' in any case. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] ROPS and latest FPC compiler (svn)
Alexandre Leclerc wrote: >> That are 1300+ revisions. Try doing a binary search. That will take 11 >> tests. > > How do a do such a thing? You know binary search right ? Use svn diff -r X, with X varying between working and newest revision, until you found X-1 working, X not. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] typed constants...
Michael Van Canneyt wrote: > Initialized constants are deprecated, and should be replaced by initialized > variables, as in Delphi: They are? Why? They have the practical advantage of being able to share them with other instances of this process. At least, if compiler+linker cooperate to put them in the right section. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] RE: Get Local IP using glib (eth0,eth1)
Jeff Pohlmeyer wrote:
>> > function to read the local IP eth0/eth1 using glibc library.
>
> This one is a little cleaner...
>
> program ip_addr;
> {$MODE OBJFPC}
>
> uses libc;
It doesn't seem like you're using libc specific functions ? You can also
write a fpc 'native' one using fpsocket and fpioctl in baseunix/unix, I
think.
Micha
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] .NET FAQ
Marco van de Voort wrote: > IMHO it would be better to start over legacy free for .NET and really target > the .NET/JVM/LLVM platform 100%. Refer them to Chrome ? ;-) Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Threads executing in sequence instead of parallel
Marco van de Voort wrote: >> Ok, maybe Sleep(0) in sysutils has to use the sched_yield syscall for >> linux. This may be a feature request to be reported in bug tracker IMHO. > > Ugly. Better define a procedure TryYieldTimeSlice() or so. No need to follow > Windowisms with archaic reasons. Yes agree, thought of this just after sending the email. The implementation might/will call Sleep(0) on the windows platform. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Threads executing in sequence instead of parallel
Graeme Geldenhuys wrote: > pointed out). This sounds like an interesting topic and something > work understanding. After all, it gave me enough headaches and a > false positive - thinking it is a bug. Btw, here is a quote from the interview, default length seems to be 150ms: (it also explains the usefulness of longer timeslices: caches) another property of SCHED_BATCH scheduling is the use of much longer timeslices. Eg. right now it's 3 seconds for a default priority SCHED_BATCH task - while normal tasks have 150 msec timeslices. For things like numeric calculations it's good to have as long timeslices as possible, to minimize the effect of cache trashing. Eg. on a sufficiently powerful CPU with a 2 MB L2 cache, the 'population time' of the cache can be as high as 10 milliseconds. So if there are two numeric calculation tasks that both fully utilize the L2 cache (in nonobvious patterns), and which context-switch every 150 milliseconds, then they will waste 10 milliseconds on cache-rebuilding in the first 6% of their timeslice. This shows up as a direct 6% slowdown of the numeric calculation jobs. Now, if SCHED_BATCH is used, and each task has a 3000 milliseconds timeslice, then the cache-rebuild overhead can be at most 0.3% - a far more acceptable number. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Threads executing in sequence instead of parallel
Graeme Geldenhuys wrote: > Ah, finally some explanation that makes sence! Thanks Micha, your > modified demo does in fact make Linux switch between threads. I also > tried the Sleep(0) you mentiond in the Sort Demo. Sleep() didn't do > the trick, but using the BurnCPU() method from Demo1 did. Ok, maybe Sleep(0) in sysutils has to use the sched_yield syscall for linux. This may be a feature request to be reported in bug tracker IMHO. > Where can I read more about the timeslices to threads under Linux. It > obviously works vastly different to Windows (and FreeBSD as someone > pointed out). This sounds like an interesting topic and something > work understanding. After all, it gave me enough headaches and a > false positive - thinking it is a bug. http://kerneltrap.org/node/517/print is an interview with one of the authors, Ingo Molnar. Search for 'timeslice length' if you don't want to read it all :-). Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Threads executing in sequence instead of parallel
Graeme Geldenhuys wrote: > The Sort Demo, which is a port of the Threads demo included with > Delphi 7, shows this clearly. I can archive and post the source, but Try inserting some Sleep(0) to force the kernel to switch to another thread ? In real worker threads you wouldn't do this of course, but here just for the visual effect. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Threads executing in sequence instead of parallel
Graeme Geldenhuys wrote:
> We have a problem!!!
>
> The test project /fcl/tests/threads.pp as well as a Sort Demo (Bubble,
> Section and Quick Sort) all execute the threads in sequence, waiting
> for the previous thread to complete, before the next one executes.
> Kind of defeats the point of using threads.
There is no bug. The time slice management is simply different. Try
attached demo (it's your demo modified). You will notice that output
will be mixed.
Linux simply gives longer timeslices to threads (for efficiency reasons).
Micha
program demo1;
{$mode objfpc}{$H+}
uses
{$IFDEF UNIX}
cthreads,
{$ENDIF}
Classes, SysUtils;
type
// counts up till 1k
TIncrementer = class(TThread)
protected
procedure Execute; override;
end;
// counts down from 1k
TDecrementer = class(TThread)
protected
procedure Execute; override;
end;
TRunThreads = class(TObject)
procedure ThreadTerminated(Sender: TObject);
private
t1, t2: TThread;
FThreadCount: integer;
public
constructor Create;
procedure RunNow;
end;
{ TRunThreads }
procedure TRunThreads.ThreadTerminated(Sender: TObject);
begin
Dec(FThreadCount);
end;
constructor TRunThreads.Create;
begin
FThreadCount := 2;
t1 := TIncrementer.Create(True);
t1.OnTerminate := @ThreadTerminated;
t1.Priority := tpLower;
t1.FreeOnTerminate := True;
t2 := TDecrementer.Create(True);
t2.OnTerminate := @ThreadTerminated;
t2.Priority := tpLower;
t2.FreeOnTerminate := True;
end;
procedure TRunThreads.RunNow;
begin
writeln('RunNow');
t1.Resume;
t2.Resume;
repeat
sleep(100);
until FThreadCount = 0;
WriteLn('All threads completed!');
end;
{ TIncrementer }
procedure BurnCPU(Multiplier: integer);
var
i, k: integer;
begin
k := 13;
for i := 0 to Multiplier*50 do
k := k div 5 * 6;
end;
procedure TIncrementer.Execute;
var
i: integer;
begin
for i := 0 to 1000 do
begin
Writeln(Classname + ': ' + IntToStr(i));
BurnCPU(3);
end;
end;
{ TDecrementer }
procedure TDecrementer.Execute;
var
i: integer;
begin
for i := 1000 downto 0 do
begin
Writeln(Classname + ': ' + IntToStr(i));
BurnCPU(4);
end;
end;
var
lRunThreads: TRunThreads;
begin
lRunThreads := TRunThreads.Create;
lRunThreads.RunNow;
writeln('Done...');
end.
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Sched_Yield, TThread, Multiplexing, Sysutils, Sleep, Win32, Linux, and Thanks
memsom wrote: >> it seemed to run in a serial fashion and not as a separate task. > > Well... Win32 at any rate, use one of the following: > > CreateSemaphore / CreateMutex / CreateEvent These are abstracted in the syncobjs unit IIRC. They are in general used to wait for something to become available and thereby putting the thread to sleep, yes. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Please Help - Need ONE HINT - Trying to get
Jason P Sage wrote: > I was trying to avoid sysutils sleep(0) in linux platform due to the size of > sysutils. The fact I don't know how to make dynamic linking work on linux > yet - (use shared libs versus static) makes my exe's bigger than I would > like. In most of my applications, size isn't an issue - but in one in > particular - it is everything. Using shared linking in linux, which I think It's better to help us increase effectiveness of smart linking ;-). > timeslice, which was my impression. I THINK that sleep refers to releasing > or yielding your entire application to other tasks the operating system is > handling. Which means its use should be minimal, and strategic - like when > you application true "should" release the CPU a bit... or is "truly" idle. I think Sleep is per-thread, not sure though. Anyway, you should never need Sleep; instead you should use mutexes or similar techniques to sleep until you need to wake up for calculation or whatever. > You see I have been playing with TThread - and I noticed that the most > currently launched thread is the one who stays running - and everything else > in my app stops until it finishes. This leads me to believe I am responsible How did you arrive at this conclusion ? It should not be so, the OS will do the scheduling for you, all threads will get their timeslice. > or some sort of mechanism. I believe TThread.Syncronize and possibly TThread > Critical Section are going to be my next area of experimentation. Critical sections are a sort of mutex indeed, about which I was talking above. > Along this lines I have changed the entire "main loop" architecture of my > bigger THREADING application - to work similar to how Lazarus makes its > executables or how a windows program works in win32 - I've built a messaging > system - that is a COMMON construct throughout all my classes, so that they > can ALL communicate in a context-less way to the main program - allowing the > main program to respond by suspending, resuming, terminating... etc any/all > of the working threads as it deems correct. Sounds like a good architecture. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Please Help - Need ONE HINT - Trying to get something like sched_yield with free pascal in linux
Jason Sage wrote: > Hi all – and thanks to who ever can respond in advance. > > > > I have a Server Like App – I need the equivalent of sched_yield Tried sleep(0); yet ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] flag or exception?
Bisma Jayadi wrote: > Hi all... > > I've always been confused about this subject... flag or exception for > error handling? Can anyone share knowledges and ideas about when and > where and why a (pascal) developer should chose one over another? Depends on taste, and they have both advantages and disadvantages. For example, it's easy to see when flag-using code is coded correctly (are the return values checked and handled?), while it's hard for exception-using code. Exceptions are usually slower as well. Exceptions are also not compatible to other languages, usually. OTOH, using exceptions is lazier/easier. If an error should never occur, then using exceptions will have more readable code, as not every function calling this must handle the possible error flag. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] DLL and Shared Libs in FPC
Jason P Sage wrote: > I have read some things here and there, but can anyone simply tell me if its > now possible to write DLL's and Linux Shared Libraries with FPC now? > > The last time I approached this topic, it was unsupported, and was still not > really something that was ready for prime time. Do you mean import/export of classes, etc. as well ? Procedural 'C-like' libraries have worked for a long time, as already stated. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Automatic objects
Marco van de Voort wrote: >> Is it right to assume they are automatically deallocated when they go >> out of scope (e.g. no need to free them manually)? > > The space of the object is cleaned. However if the object uses dyn memory, > you will have to call the destructor (destroy) to make the object cleanup. I think in theory it's possible for the compiler to automatically call the destructor always when leaving the function (think implicit exception handlers here), but I guess not so many people are using them anymore in new code that it would be worth it. Would need a new keyword or something too, as it's not backward-compatible obviously. Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] CRT (Console Unit) Problems with Fedora 5
Jason P Sage wrote: > A good example is that FreeVision and the FPC IDE do not work on my Linux > either. Elaborate on "do not work" please ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Common OpenMP syntax?
Michael Van Canneyt wrote: >> The latter. > > Which is why I think that it's better to have them as local functions, > instead of having to introduce a lot of new functions. There is no real reason to restrict 'parallel' to local functions, is there ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Common OpenMP syntax?
Vinzent Hoefler wrote: > On Wednesday 26 July 2006 09:07, Micha Nelissen wrote: >> How many copies ? > > Omp.Get_Num_Threads(), AFAICS. Ah the number of threads is determined by the RTL, and any parallel block must be written flexible, so that it can work for any given number of threads ? Micha ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
