I spotted a bug in current fpc 1.9.1 and the Round function. I was not able to reproduce…
Trying to make latest cvs on Win98 , I receive a "windres: no resources" compile error - any ideas ?
Mailing lists / lazarus / 2003-11
Win32 Compile errors
5 messages · 4 people ·
- Mattias Gaertner 2
- C Western 1
- Florian Klaempfl 1
- Steve Gatenby 1
Reply by email Subscribe to this list Your subscription
Read-only archive. Use j and k to move between messages.
I spotted a bug in current fpc 1.9.1 and the Round function. I was not able to reproduce…
Mattias Gaertner wrote 16 lines
Hi,
I spotted a bug in current fpc 1.9.1 and the Round function.
I was not able to reproduce the error in a small context or find the bug in
the fpc bug list. Perhaps the needed context has something to do with the
gtk libs. So, maybe someone else knows this bug:
The bug appears in gtkwinapi.inc 7773 in the line "Result:=Round(Value);":
Program received signal SIGFPE, Arithmetic exception.
0x0813ca57 in fpc_round ()
Value is 0, so rounding should be pretty easy.
Adding a writeln(Value); in front "fixes" the bug.
Any clues/hints?
Can you set a break point before the round call and post an fpu dump (including status registers)?
I spotted a bug in current fpc 1.9.1 and the Round function. I was not able to reproduce…
Mattias Gaertner wrote 16 lines
>>I spotted a bug in current fpc 1.9.1 and the Round function.
>>I was not able to reproduce the error in a small context or find the bug
>>in the fpc bug list. Perhaps the needed context has something to do with
>>the gtk libs. So, maybe someone else knows this bug:
>>
>>The bug appears in gtkwinapi.inc 7773 in the line
>>"Result:=Round(Value);":
>>
>>Program received signal SIGFPE, Arithmetic exception.
>>0x0813ca57 in fpc_round ()
>>
>>Value is 0, so rounding should be pretty easy.
>>Adding a writeln(Value); in front "fixes" the bug.
>>
>>
>>
I have had this sort of problem (though not in your specific context). In my case it was caused by a problem elsewhere - either: a) The floating point stack is full
b) A fpu exception has been raised proviously, but is discovered in the round function when exceptions are unmasked
Colin
I spotted a bug in current fpc 1.9.1 and the Round function. I was not able to reproduce…
Show quoted text 22 lines
Mattias Gaertner wrote:
> Hi,
>
> I spotted a bug in current fpc 1.9.1 and the Round function.
> I was not able to reproduce the error in a small context or find the bug
> in the fpc bug list. Perhaps the needed context has something to do with
> the gtk libs. So, maybe someone else knows this bug:
>
> The bug appears in gtkwinapi.inc 7773 in the line
> "Result:=Round(Value);":
>
> Program received signal SIGFPE, Arithmetic exception.
> 0x0813ca57 in fpc_round ()
>
> Value is 0, so rounding should be pretty easy.
> Adding a writeln(Value); in front "fixes" the bug.
>
> Any clues/hints?
Can you set a break point before the round call and post an fpu dump
(including status registers)?
Breakpoint 1, TGTKOBJECT__SETSCROLLINFO (HANDLE=144650600, SBSTYLE=1, SCROLLINFO=
{CBSIZE = 28, FMASK = 103, NMIN = 0, NMAX = 0, NPAGE = 244, NPOS = 0, NTRACKPOS = 0}, BREDRAW=false,
this=$81a3594, result=0, ADJUSTMENT=$89f5e18, SCROLL=$89f5158, NEWPOLICY=-1073744992) at gtkwinapi.inc:7776
7776 gtkwinapi.inc: No such file or directory.
in gtkwinapi.inc
Current language: auto; currently pascal
(gdb) info float
R7: Empty 0x00000000000000000000
=>R6: Empty 0x40058bffffffffffe000
R5: Empty 0x3ffed406483ed274388a
R4: Empty 0x4000c000000000000000
R3: Empty 0x4006f300000000000000
R2: Empty 0x3ffe95a578c9c6fc07ee
R1: Empty 0x3ffe95a578c9c6fc07ee
R0: Empty 0x4006b400000000000000
Status Word: $3061 IE PE SF TOP: 6
Control Word: $037f IM DM ZM OM UM PM
PC: Extended Precision (64-bits)
RC: Round to nearest
Tag Word: $ffff
Instruction Pointer: $00:$00000000
Operand Pointer: $00:$00000000
Opcode: $0000
(gdb) c
Continuing.
Program received signal SIGFPE, Arithmetic exception. $0813ca47 in fpc_round ()
Mattias
Hi,
I spotted a bug in current fpc 1.9.1 and the Round function. I was not able to reproduce the error in a small context or find the bug in the fpc bug list. Perhaps the needed context has something to do with the gtk libs. So, maybe someone else knows this bug:
The bug appears in gtkwinapi.inc 7773 in the line "Result:=Round(Value);":Program received signal SIGFPE, Arithmetic exception. 0x0813ca57 in fpc_round ()
Value is 0, so rounding should be pretty easy.
Adding a writeln(Value); in front "fixes" the bug.
Any clues/hints?
Mattias