Re: [fpc-pascal] socket timeout
Will this be a CPU hog: t := Now; while fprecv(s, @buf, buf_len, MSG_DONTWAIT) < 0 do begin if Now - t > 0.1 * timeout then Break; // approx. seconds end; Thanks 2013/4/16 Luca Olivetti > Al 16/04/13 12:20, En/na Michael Schnell ha escrit: > > On 04/16/2013 11:26 AM, Xiangrong Fang wrote: > >> How can I set the connection and read/write timeout when using socket > >> in free pascal? > > I seem to remember that synapse can do this. > > Only in svn and it does it this way > > if FConnectionTimeout > 0 then > begin > // connect in non-blocking mode > b := NonBlockMode; > NonBlockMode := true; > SockCheck(synsock.Connect(FSocket, Sin)); > if (FLastError = WSAEINPROGRESS) OR (FLastError = WSAEWOULDBLOCK) then > if not CanWrite(FConnectionTimeout) then > FLastError := WSAETIMEDOUT; > NonBlockMode := b; > end > else > SockCheck(synsock.Connect(FSocket, Sin)); > > > NonBlockMode in turn uses fpIoCtl with the FIONBIO parameter, > WSAEINPROGRESS=EsysEINPROGRESS, WSAEWOULDBLOCK=EsysEWOULDBLOCK and > CanWrite does an fpSelect > > Bye > -- > Luca > > ___ > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] socket timeout
On Wednesday 17 April 2013 09:17:09 you wrote: > Will this be a CPU hog: > > t := Now; > while fprecv(s, @buf, buf_len, MSG_DONTWAIT) < 0 do begin > if Now - t > 0.1 * timeout then Break; // approx. seconds > end; > > Thanks I did not read the whole thread, but that you favorite blocking mode (which I don't use). For MSG_DONTWAIT as far I know you have to change to the nonblocking mode. > > > 2013/4/16 Luca Olivetti > > > Al 16/04/13 12:20, En/na Michael Schnell ha escrit: > > > On 04/16/2013 11:26 AM, Xiangrong Fang wrote: > > >> How can I set the connection and read/write timeout when using socket > > >> in free pascal? > > > > > > I seem to remember that synapse can do this. > > > > Only in svn and it does it this way > > > > if FConnectionTimeout > 0 then > > begin > > > > // connect in non-blocking mode > > b := NonBlockMode; > > NonBlockMode := true; > > SockCheck(synsock.Connect(FSocket, Sin)); > > if (FLastError = WSAEINPROGRESS) OR (FLastError = WSAEWOULDBLOCK) then > > > > if not CanWrite(FConnectionTimeout) then > > > > FLastError := WSAETIMEDOUT; > > > > NonBlockMode := b; > > > > end > > else > > > > SockCheck(synsock.Connect(FSocket, Sin)); > > > > NonBlockMode in turn uses fpIoCtl with the FIONBIO parameter, > > WSAEINPROGRESS=EsysEINPROGRESS, WSAEWOULDBLOCK=EsysEWOULDBLOCK and > > CanWrite does an fpSelect > > > > Bye > > -- > > Luca > > > > ___ > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] socket timeout
I thought MSG_DONTWAIT is actually the flag to switch to nonblocking mode, isn't it? As a matter of fact, the fprecv() above would get a result of -1 and a sockterror of "Try Again", which is exactly the behavior of nonblocking read. 2013/4/17 Rainer Stratmann > On Wednesday 17 April 2013 09:17:09 you wrote: > > Will this be a CPU hog: > > > > t := Now; > > while fprecv(s, @buf, buf_len, MSG_DONTWAIT) < 0 do begin > > if Now - t > 0.1 * timeout then Break; // approx. > seconds > > end; > > > > Thanks > > I did not read the whole thread, but that you favorite blocking mode > (which I > don't use). > > For MSG_DONTWAIT as far I know you have to change to the nonblocking mode. > > > > > > > 2013/4/16 Luca Olivetti > > > > > Al 16/04/13 12:20, En/na Michael Schnell ha escrit: > > > > On 04/16/2013 11:26 AM, Xiangrong Fang wrote: > > > >> How can I set the connection and read/write timeout when using > socket > > > >> in free pascal? > > > > > > > > I seem to remember that synapse can do this. > > > > > > Only in svn and it does it this way > > > > > > if FConnectionTimeout > 0 then > > > begin > > > > > > // connect in non-blocking mode > > > b := NonBlockMode; > > > NonBlockMode := true; > > > SockCheck(synsock.Connect(FSocket, Sin)); > > > if (FLastError = WSAEINPROGRESS) OR (FLastError = WSAEWOULDBLOCK) > then > > > > > > if not CanWrite(FConnectionTimeout) then > > > > > > FLastError := WSAETIMEDOUT; > > > > > > NonBlockMode := b; > > > > > > end > > > else > > > > > > SockCheck(synsock.Connect(FSocket, Sin)); > > > > > > NonBlockMode in turn uses fpIoCtl with the FIONBIO parameter, > > > WSAEINPROGRESS=EsysEINPROGRESS, WSAEWOULDBLOCK=EsysEWOULDBLOCK and > > > CanWrite does an fpSelect > > > > > > Bye > > > -- > > > Luca > > > > > > ___ > > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > ___ > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Same code from LCL project throw error in console application
El 16/04/2013 22:46, Krzysztof escribió: Thanks! exInvalidOp exception mask solved problem. But I don't understand. Same error I had in fpGUI project. fpGUI interface doesn't use GTK, QT etc, it drawing over pure X11 so how it is possible that application catch GTK errors? Hello, Almost any 'C' library mask that errors, so the programmers do not take them in account and anyone, like the X11, can raise that exceptions. -- ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Re: Correct way for partial generic specialization
> You're missing the again. :) *self-headcut > Not currently. It might be a useful extension however. I'll need to test > what Delphi says about this... OK, I'll wait -- View this message in context: http://free-pascal-general.1045716.n5.nabble.com/Correct-way-for-partial-generic-specialization-tp5714160p5714175.html Sent from the Free Pascal - General mailing list archive at Nabble.com. ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: Correct way for partial generic specialization
Am 17.04.2013 11:11, schrieb leledumbo: You're missing the again. :) *self-headcut Can it be that your E-Mail client misinterprets the "<...>" (with only one string as "value") as HTML tags? Because I can't see the "" I had written to you in your quotation. Not currently. It might be a useful extension however. I'll need to test what Delphi says about this... OK, I'll wait So... Delphi does not support that, but I still think that this is a valid extension of generics and can be used for all generic-able types instead of only classes. Regards, Sven ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Fpcup for FreeBSD: feedback welcome
Just uploaded a new version of fpcup (automatic FPC+Lazarus svn downloader/installer for both stable and trunk) at https://bitbucket.org/reiniero/fpcup/downloads I welcome feedback from FreeBSD users - the binary should work on FreeBSD 9 x64... For the Lazarus installation part, it picks up newer/ports gdb installs in /usr/local/bin Prerequisites: - none for PC-BSD; probably none for FreeBSD - optional but recommended: newer gdb, e.g. installed via ports tree - if running Lazarus: GTK environment Thanks, Reinier ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] socket timeout
On Wednesday 17 April 2013 10:22:58 you wrote: > I thought MSG_DONTWAIT is actually the flag to switch to nonblocking mode, > isn't it? May be. > As a matter of fact, the fprecv() above would get a result of -1 > and a sockterror of "Try Again", which is exactly the behavior of > nonblocking read. Yes it seems so. I do always separately switch to the nonblocking mode before. > > > 2013/4/17 Rainer Stratmann > > > On Wednesday 17 April 2013 09:17:09 you wrote: > > > Will this be a CPU hog: > > > > > > t := Now; > > > while fprecv(s, @buf, buf_len, MSG_DONTWAIT) < 0 do begin > > > > > > if Now - t > 0.1 * timeout then Break; // approx. > > > > seconds > > > > > end; > > > > > > Thanks > > > > I did not read the whole thread, but that you favorite blocking mode > > (which I > > don't use). > > > > For MSG_DONTWAIT as far I know you have to change to the nonblocking > > mode. > > > > > 2013/4/16 Luca Olivetti > > > > > > > Al 16/04/13 12:20, En/na Michael Schnell ha escrit: > > > > > On 04/16/2013 11:26 AM, Xiangrong Fang wrote: > > > > >> How can I set the connection and read/write timeout when using > > > > socket > > > > > > >> in free pascal? > > > > > > > > > > I seem to remember that synapse can do this. > > > > > > > > Only in svn and it does it this way > > > > > > > > if FConnectionTimeout > 0 then > > > > begin > > > > > > > > // connect in non-blocking mode > > > > b := NonBlockMode; > > > > NonBlockMode := true; > > > > SockCheck(synsock.Connect(FSocket, Sin)); > > > > if (FLastError = WSAEINPROGRESS) OR (FLastError = WSAEWOULDBLOCK) > > > > then > > > > > > if not CanWrite(FConnectionTimeout) then > > > > > > > > FLastError := WSAETIMEDOUT; > > > > > > > > NonBlockMode := b; > > > > > > > > end > > > > else > > > > > > > > SockCheck(synsock.Connect(FSocket, Sin)); > > > > > > > > NonBlockMode in turn uses fpIoCtl with the FIONBIO parameter, > > > > WSAEINPROGRESS=EsysEINPROGRESS, WSAEWOULDBLOCK=EsysEWOULDBLOCK and > > > > CanWrite does an fpSelect > > > > > > > > Bye > > > > -- > > > > Luca > > > > > > > > ___ > > > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > > > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > > > > ___ > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] socket timeout
For connect, I wrote a unit that implements a connect timeout, see:
http://sourceforge.net/p/pascalscada/code/HEAD/tree/trunk/src/scada/sockets_w32_w64.pas#l70
or
http://sourceforge.net/p/pascalscada/code/HEAD/tree/trunk/src/scada/sockets_unix.pas#l48
If you want work in blocking mode after connect, do something like this:
fsocket = {$IFDEF WINDOWS}Socket{$ELSE}fpSocket{$ENDIF}(PF_INET,
SOCK_STREAM, IPPROTO_TCP)
//set socket options.
setblockingmode(fsocket,MODE_NONBLOCKING);
if connect_with_timeout(focket,@channel,sizeof(channel),FTimeoutInMs)<>0
then begin
//check errors
end;
setblockingmode(FSocket,MODE_BLOCKING);
2013/4/17 Rainer Stratmann :
> On Wednesday 17 April 2013 10:22:58 you wrote:
>> I thought MSG_DONTWAIT is actually the flag to switch to nonblocking mode,
>> isn't it?
> May be.
>> As a matter of fact, the fprecv() above would get a result of -1
>> and a sockterror of "Try Again", which is exactly the behavior of
>> nonblocking read.
> Yes it seems so.
>
> I do always separately switch to the nonblocking mode before.
>>
>>
>> 2013/4/17 Rainer Stratmann
>>
>> > On Wednesday 17 April 2013 09:17:09 you wrote:
>> > > Will this be a CPU hog:
>> > >
>> > > t := Now;
>> > > while fprecv(s, @buf, buf_len, MSG_DONTWAIT) < 0 do begin
>> > >
>> > > if Now - t > 0.1 * timeout then Break; // approx.
>> >
>> > seconds
>> >
>> > > end;
>> > >
>> > > Thanks
>> >
>> > I did not read the whole thread, but that you favorite blocking mode
>> > (which I
>> > don't use).
>> >
>> > For MSG_DONTWAIT as far I know you have to change to the nonblocking
>> > mode.
>> >
>> > > 2013/4/16 Luca Olivetti
>> > >
>> > > > Al 16/04/13 12:20, En/na Michael Schnell ha escrit:
>> > > > > On 04/16/2013 11:26 AM, Xiangrong Fang wrote:
>> > > > >> How can I set the connection and read/write timeout when using
>> >
>> > socket
>> >
>> > > > >> in free pascal?
>> > > > >
>> > > > > I seem to remember that synapse can do this.
>> > > >
>> > > > Only in svn and it does it this way
>> > > >
>> > > > if FConnectionTimeout > 0 then
>> > > > begin
>> > > >
>> > > > // connect in non-blocking mode
>> > > > b := NonBlockMode;
>> > > > NonBlockMode := true;
>> > > > SockCheck(synsock.Connect(FSocket, Sin));
>> > > > if (FLastError = WSAEINPROGRESS) OR (FLastError = WSAEWOULDBLOCK)
>> >
>> > then
>> >
>> > > > if not CanWrite(FConnectionTimeout) then
>> > > >
>> > > > FLastError := WSAETIMEDOUT;
>> > > >
>> > > > NonBlockMode := b;
>> > > >
>> > > > end
>> > > > else
>> > > >
>> > > > SockCheck(synsock.Connect(FSocket, Sin));
>> > > >
>> > > > NonBlockMode in turn uses fpIoCtl with the FIONBIO parameter,
>> > > > WSAEINPROGRESS=EsysEINPROGRESS, WSAEWOULDBLOCK=EsysEWOULDBLOCK and
>> > > > CanWrite does an fpSelect
>> > > >
>> > > > Bye
>> > > > --
>> > > > Luca
>> > > >
>> > > > ___
>> > > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org
>> > > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal
>> >
>> > ___
>> > fpc-pascal maillist - fpc-pascal@lists.freepascal.org
>> > http://lists.freepascal.org/mailman/listinfo/fpc-pascal
> ___
> fpc-pascal maillist - fpc-pascal@lists.freepascal.org
> http://lists.freepascal.org/mailman/listinfo/fpc-pascal
--
The best regards,
Fabio Luis Girardi
PascalSCADA Project
http://sourceforge.net/projects/pascalscada
http://www.pascalscada.com
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] How can retrive value from OUT parameter?
Hi All, I have an application build with Delphi7+ADO connecting to MSSQL database, and planning to migrate to FPC+SQLdb. It has many stored procedures which have OUT parameters to retrieve value from database. I am trying to migrate by using TMSSQLConnection+TSQLQuery, but still I can not settle it. I can execute and pass values to stored procedures, but not the other way. The RETURN_VALUE (default result) can not be retrieved either. I would appreciate any advice. Toru ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Bending System.IndexByte to my will!!?
Hi all,
(FPC 2.6.2 x86, trunk x64, on Windows; found similar behaviour on Linux x64)
Busy getting DBase III memo support correct ;)
In fcl-db\src\dbase\dbf_common.pas we find this part used to e.g. find
end of file ($1A/ASCII 26) markers in a buffer with memo data: [1]
The problem is: the first MemScan function doesn't work:
System.IndexByte just always return -1 even if the character is there.
Tried system.indexbyte with a simple test program and of course it worked.
If I comment out $ifdef fpc to force use of the second function, the $1A
is found perfectly.
Where lies the problem?
Thanks,
Reinier
[1]
{$ifdef FPC}
function MemScan(const Buffer: Pointer; Chr: Byte; Length: Integer):
Pointer;
var
I: Integer;
begin
I := System.IndexByte(Buffer, Length, Chr);
if I = -1 then
Result := nil
else
Result := Buffer+I;
end;
{$else}
function MemScan(const Buffer: Pointer; Chr: Byte; Length: Integer):
Pointer;
asm
PUSHEDI
MOV EDI,Buffer
MOV AL, Chr
MOV ECX,Length
REPNE SCASB
MOV EAX,0
JNE @@1
MOV EAX,EDI
DEC EAX
@@1:POP EDI
end;
{$endif}
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Bending System.IndexByte to my will!!?
Den 18-04-2013 07:57, Reinier Olislagers skrev:
Hi all,
(FPC 2.6.2 x86, trunk x64, on Windows; found similar behaviour on Linux x64)
Busy getting DBase III memo support correct ;)
In fcl-db\src\dbase\dbf_common.pas we find this part used to e.g. find
end of file ($1A/ASCII 26) markers in a buffer with memo data: [1]
The problem is: the first MemScan function doesn't work:
System.IndexByte just always return -1 even if the character is there.
Tried system.indexbyte with a simple test program and of course it worked.
If I comment out $ifdef fpc to force use of the second function, the $1A
is found perfectly.
Where lies the problem?
Thanks,
Reinier
[1]
{$ifdef FPC}
function MemScan(const Buffer: Pointer; Chr: Byte; Length: Integer):
Pointer;
var
I: Integer;
begin
I := System.IndexByte(Buffer, Length, Chr);
if I = -1 then
Result := nil
else
Result := Buffer+I;
end;
{$else}
function MemScan(const Buffer: Pointer; Chr: Byte; Length: Integer):
Pointer;
asm
PUSHEDI
MOV EDI,Buffer
MOV AL, Chr
MOV ECX,Length
REPNE SCASB
MOV EAX,0
JNE @@1
MOV EAX,EDI
DEC EAX
@@1:POP EDI
end;
{$endif}
Since you have a typed const pointer instead of an untyped const
parameter, you need to do this:
System.IndexByte(PByte(Buffer)^, Length, Chr);
___
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Bending System.IndexByte to my will!!?
On 18-4-2013 8:14, Jeppe Græsdal Johansen wrote: > Den 18-04-2013 07:57, Reinier Olislagers skrev: >> In fcl-db\src\dbase\dbf_common.pas we find this part used to e.g. find >> end of file ($1A/ASCII 26) markers in a buffer with memo data: [1] >> >> The problem is: the first MemScan function doesn't work: >> System.IndexByte just always return -1 even if the character is there. >> Tried system.indexbyte with a simple test program and of course it >> worked. >> If I comment out $ifdef fpc to force use of the second function, the $1A >> is found perfectly. >> >> Where lies the problem? > Since you have a typed const pointer instead of an untyped const > parameter, you need to do this: > > System.IndexByte(PByte(Buffer)^, Length, Chr); > Thanks a lot, that works beautifully! ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] How to use TThread?
First of all change the order of lines 25-26 inherited Create(False); FreeOnTerminate := True; instead of FreeOnTerminate := True; inherited Create(False); The, drop the DoExecute procedure and replace it by the Execute procedure. It is useless to Synchronize a thread's procedure inside itself ! After the creation, the thread executes immediately the overriden procedure Execute because, at creation time (inherited Create(False); ) the thread is created in the state NON SUSPENDED. The use of the DoExecute is obviously useless our program will execute the procedure Excute only once as there is no loop. Antonio Fortuny Senior Software engineer 220, avenue de la Liberté L-4602 Niederkorn Tel.: +352 58 00 93 - 93 www.sitasoftware.lu Le 18/04/2013 07:03, silvioprog a écrit : Hi, I tried to use TThread in a small sample, but it don't works. The code is very simple, and I'm sending it in attached. Thanks, -- Silvio Clécio My public projects - github.com/silvioprog ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
