[Oberon] PO2013 - Show Heap Blocks
Jörg
joerg.straube at iaeth.ch
Fri Jun 19 17:43:57 CEST 2020
John
In case you're interested, here the code:
PROCEDURE printf*(s: ARRAY OF CHAR; i1, i2: INTEGER; str: ARRAY OF CHAR);
(* supported format %[length]type, type = d, i, x or s, and escape characters \n, \t *)
VAR n, j, len, par, x: INTEGER; ch, nxCh: CHAR; imm: BOOLEAN;
BEGIN
imm := immediate; immediate := FALSE;
par := 1; n := 0; ch := s[0];
WHILE (ch # 0X) & (n < LEN(s)) DO
j := n+1; nxCh := s[j];
IF ch = "%" THEN (* handle format *)
len := 0;
WHILE ("0" <= nxCh) & (nxCh <= "9") DO len := len * 10 + ORD(nxCh) - ORD("0"); INC(j); nxCh := s[j] END;
IF nxCh = "s" THEN String(str); INC(j)
ELSIF (nxCh # "d") & (nxCh # "i") & (nxCh # "x") THEN Char("%")
ELSE
x := i2; IF par = 1 THEN x := i1 END; INC(par);
IF nxCh = "x" THEN Hex(x) ELSE (* i, d *) Int(x, len) END;
INC(j)
END
ELSIF ch = "\" THEN (* handle escape characters *)
IF nxCh = "n" THEN Ln; INC(j)
ELSIF nxCh = "t" THEN Char(9X); INC(j)
ELSE Char("\")
END
ELSE
Char(ch)
END;
n := j; ch := s[n]
END;
immediate := imm
END printf;
Am 19.06.20, 17:23 schrieb "Joerg" <joerg.straube at iaeth.ch>:
John
Seems like variadic but it is not.
printf() has fix 4 arguments as I found out that this is enough for most cases. the first is the string with formatting commands, then two INTEGERs and one string (for a filename or so)
If after a % there is a number, d, h or s it’s interpreted as argument else it’s just a %.
br
Jörg
> Am 19.06.2020 um 16:56 schrieb John R. Strohm <strohm at airmail.net>:
>
> Out.printf(), with VARIADIC arguments, in Oberon???
>
> And that %d%) bothers me, for some reason...
>
> --- joerg.straube at iaeth.ch wrote:
>
> From: Jörg <joerg.straube at iaeth.ch>
> To: ETH Oberon and related systems <oberon at lists.inf.ethz.ch>
> Subject: Re: [Oberon] PO2013 - Show Heap Blocks
> Date: Fri, 19 Jun 2020 10:37:08 +0200
>
> Tomas
>
> The following line will give you the stats as well:
> Out.printf(“Heap space used: %d bytes (=%d%)\n“,
> Kernel.allocated,
> Kernel.allocated * 100 DIV (Kernel.heapLim - Kernel.heapOrg), ““);
>
> That‘s basically what System.Watch does.
>
> br
> Jörg
>
>
>> Am 19.06.2020 um 10:12 schrieb Tomas Kral <thomas.kral at email.cz>:
>>
>> Hi,
>>
>> This is my understanding so far, it stats the heap.
>> It also shows free block last to the limit, because heap ends 16 bytes
>> earlier, so is not MOD 32.
>>
>> MODULE Heap; (*TK 14.6.2020 show heap blocks*)
>> IMPORT S := SYSTEM, Kernel, Out;
>>
>> VAR heapOrg, MemLim, heapLim: INTEGER;
>>
>> PROCEDURE Scan*;
>> VAR p, q, mark, tag, size, used, free: LONGINT;
>> BEGIN p := heapOrg; used := 0; free := 0;
>> Out.Hex(heapOrg); Out.Ln; (*60000H 4-bit colour display*)
>> Out.Hex(heapLim); Out.Ln; (*9FEF0H*)
>> REPEAT S.GET(p+4, mark); q := p; (*mark: 0 used, -1 freed, >0 marked*)
>> IF mark = -1 THEN S.GET(p, size)
>> ELSE S.GET(p, tag); S.GET(tag, size); (*size := p - q*) END ;
>> IF (size MOD 32 # 0) & (size MOD 64 # 0) & (size MOD 128 # 0) & (size MOD 256 # 0) THEN
>> Out.Hex(p); Out.Hex(size); Out.Hex(mark); Out.Ln (*free block to 9FEF0H*)
>> END ;
>> IF mark > 0 THEN INC(used, size) (*none after kernel scan*)
>> ELSIF mark = 0 THEN INC(used, size)
>> ELSIF mark = -1 THEN INC(free, size) END ;
>> INC(p, size)
>> UNTIL p >= heapLim(*-30000H*) ;
>> Out.Int(used, 8); (*value*) Out.Int(used*100 DIV free, 3); (*percentage*)
>> Out.Hex(free+used); Out.Ln (*3FEF0 = 9FEF0H-60000H, or less if fragmented*)
>> END Scan;
>>
>> PROCEDURE Lists*;
>> VAR p, list0, list1, list2, list3: INTEGER;
>> BEGIN
>> p := S.ADR(Kernel.MemLim);
>> S.GET(p+8, list0);
>> S.GET(p+12, list1);
>> S.GET(p+16, list2);
>> S.GET(p+20, list3);
>> Out.Hex(list3); Out.Hex(list2); Out.Hex(list1); Out.Hex(list0); Out.Ln
>> END Lists;
>>
>> BEGIN S.GET(12, MemLim); S.GET(24, heapOrg); heapLim := MemLim
>> END Heap.
>>
>> Heap.Scan
>> Heap.Lists
>>
>>
>> --
>> Tomas Kral <thomas.kral at email.cz>
>> --
>> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
>> https://lists.inf.ethz.ch/mailman/listinfo/oberon
>
> --
> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
> https://lists.inf.ethz.ch/mailman/listinfo/oberon
>
>
> --
> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
> https://lists.inf.ethz.ch/mailman/listinfo/oberon
More information about the Oberon
mailing list