Sign in Join

Live chat for Free Pascal, Lazarus, Delphi and every Pascal programmer.

Mailing lists / fpc-pascal / 2002-05

Using FPC 1.1 for bootstraping compiler v1.0.7

5 messages · 4 people · –

Reply by email Subscribe to this list Your subscription

Read-only archive. Use j and k to move between messages.

Sergey Korshunoff

Hi!

There was a situation when I only have 1.1 binaries. FPC v1.1 do not want to compile compiler v1.0.7 without a patch. Patch is mainly do
$FFFFFFFF ==> -1
replacing (v1.1 complain about outrange longint initialization). Is there in FPC option to setup for such case a behavior as in v1.0.7 ?

Regards,
Sergey Korshunoff

PS: if you remove
override RELEASE=1
from root Makefile.fpc, then you can not build a 1.1 dist with 1.1 compiler :(((

There is no such problem for v1.0.7 ( I work under RH Linux v7.2)

Jonas Maebe
to Sergey Korshunoff
Sergey Korshunoff wrote 6 lines
There was a situation when I only have 1.1 binaries. FPC v1.1 do not want to compile compiler v1.0.7 without a patch. Patch is mainly do $FFFFFFFF ==> -1 replacing (v1.1 complain about outrange longint initialization). Is there in FPC option to setup for such case a behavior as in v1.0.7 ?

No. This requires 64bit constant support, which 1.0.7 doesn't have.

Show quoted text 5 lines
PS: if you remove override RELEASE=1 from root Makefile.fpc, then you can not build a 1.1 dist with 1.1 compiler :(((

What do you mean with "dist"? A collection of zipiles to distribute? If so, why would you want to remove the "dist=1"?

Jonas

Sergey Korshunoff
to Jonas Maebe
Show quoted text 9 lines
On vrijdag, mei 31, 2002, at 01:45 , Sergey Korshunoff wrote: > There was a situation when I only have 1.1 binaries. > FPC v1.1 do not want to compile compiler v1.0.7 without a patch. > Patch is mainly do > $FFFFFFFF ==> -1 > replacing (v1.1 complain about outrange longint initialization). > Is there in FPC option to setup for such case a behavior as in v1.0.7 ? No. This requires 64bit constant support, which 1.0.7 doesn't have.

I patch a sources of v1.0.7 compiler to allow compile that's with fpc 1.1. Question is: is there in FPC 1.1 a flag to allow behave as Kylix in

      a32bitInteger:=$FFFFFFFF;

case? (Kylix only give warning in such case, FPC v1.1 aborts compilation)

Show quoted text 7 lines
> PS: if you remove > override RELEASE=1 > from root Makefile.fpc, then you can not build a 1.1 dist with 1.1 > compiler What do you mean with "dist"? A collection of zipiles to distribute? If so, why would you want to remove the "dist=1"?

I use
make rpm
to build FPC packet under Linux (RPM allow an easy remove). And I want to build UNITS with debug info (not optimized version). But this is not possible: FPC v1.1 can not build full tree without optimization.

Regards,
Sergey Korshunoff

Julian Calderon Almendros
to Sergey Korshunoff

If I try:
...

	CONST
		QWord_Var : QWord = $FFFFFFFFFFFFFFFF;
            ...

It has the next error: Invalid Expression. In others ocassions FPC interpretes that the constant is a real 80 or so. However if I try:

            PROGRAM ImTestingQWord;
	VAR
		QWord_Var : QWord;
	BEGIN
		ReadLn(QWord_Var);
		WriteLn(QWord_Var)
	END.

And then I execute ImTestingQWord:
c>$FFFFFFFFFFFFFFFF
c>18446744073709551615
????
And too there are problems with QWord and the case statment.

Is it a bug?

THANKS.

Florian Klaempfl
to Julian Calderon Almendros
you wrote 21 lines
If I try: ... CONST QWord_Var : QWord = $FFFFFFFFFFFFFFFF; ... It has the next error: Invalid Expression. In others ocassions FPC interpretes that the constant is a real 80 or so. However if I try: PROGRAM ImTestingQWord; VAR QWord_Var : QWord; BEGIN ReadLn(QWord_Var); WriteLn(QWord_Var) END. And then I execute ImTestingQWord: c>$FFFFFFFFFFFFFFFF c>18446744073709551615 ???? And too there are problems with QWord and the case statment. Is it a bug?

Not really :) , I guess you're using 1.0.x. The QWord and Int64 type aren't fully implemented in these compiler version. Only 1.1 supports the constructs mentioned above.