Google
 

Trailing-Edge - PDP-10 Archives - decuslib20-01 - decus/20-0002/uorfix.bug
There is 1 other file named uorfix.bug in the archive. Click here to see a list.
\UR#1\ PRELOAD (ARRAYS) WAS ALLOWING NON-INTEGER REPLICATION
      COUNT. FIX IS IN ARRAY/11
\UR#2\ UOR USES RAID AS DEBUGGER LOADED WITH SAILDD
      PARSE/16,18,19,21 (7 FIXES) GOGOL/30
\UR#3\ BIG LEAP
      HEAD (MANY PLACES), LEPRUN (MANY)
\UR#4\ ALLOW := FOR _, >= FOR GEQ ETC
      SYM/5
\UR#5\ BETTER LISTING FOR CONDITIONAL ASSYMBLY
	GEN/24
\UR#6\ VERIFY!DATUM. LEAP/13
	GEN/4/19
\UR#7\ REQUIRE OVERLAP!OK.
	GEN/16/46, LEAP, GOGOL/8

\UR#8\ IN SAILDD PUT A LINEFEED AFTER DISPLAYED LINE.
	COMSER/8
\UR#9\ CVSTR - MAKE SURE BIT 35 OFF. SCREWS UP COMPILER IF
	CALLED AT COMPILE TIME OTHERWISE.
	CVSTR/17
\UR#10\ ADD CATLST,COPARR TO SAIDM1 SO WON'T NEED LEAP FOR
        CONTEXTS. ARYSER/4
\UR#11\ GIVE BETTER MESSAGE FOR MISSING STEP AMOUNT (FOR STMT)
	HEL, STATS
\UR#12\ ITEMVAR FORMAL NEEDS ITEMVAR ACTUAL. BUGFIX IN CODE
	DOESN'T WORK FOR ARRYIN,ARROUT,ARRBLT.
	STATS/35
\UR#13\ IN!CONTEXT. GET RID OF INFINITE LOOP. HANDLE STRING ELEMENTS
	BY RETURNING POINTER WITH LEFT HALF ALL ONES.
\UR#14\ DECLARATION(RESERVED) WAS LEAVING NOMACR ON
        GEN/22
\UR#15\ CONTEXT:ARRAY[N] (CONELM IN WRDGET) BUGGY
        WRDGET/15
\UR#16\ JRL - 6/7/77 BUG IN FORGET LSTARR FROM CONTEXT
	WRDGET/5
\UR#17\ PDR - 3/78 Bug in line labelled CS4 in HEL: I removed
	an unneeded call on OFFDLM ... was popping one too many times
	to cause a BPOP error on selection of the last case in a casec.
	HEL/16-2
\UR#18\ PDR - 3/78 If the largest case label appeared first in a list
	of them, it was ignored. STATS, Pg. 17-2, just before CASEME
\UR#19\ PDR - 4/3/78 If first thing in a case body was the ELSE clause,
	a BPOP error occurred.
	Removed extraneous pop sequence at CASEME+10 in STATS
\UR#20\ PDR 4/26/78 (3 parts) fixed xcall's access to the AC for 
	error msg
		1 of 3: CASSTR:+10,STATS/16
		2 of 3: STATS, just before CLD1:, pg 18-2
		3 of 3: after CASE1:, STATS/17
\UR#21\ JRL late 1977 Bug in Leap's HASH macro caused ILL MEM REF
        with a shared runtime system. LEPRUN pg. 8
\UR#22\ JRL 4/28/78 Erroneous error message in FORGET about missing
        descriptor. WRDGET pg. 5

\UR#23\ JRL 5/2/78. OUT(CHAN,STRING) did not maintain buffer byte
        count correctly (always said there was one more byte available
        than there really was). OUT, itself, compensated for this but
        WORDOUT (used as BYTE output on channel opened in an ASCII mode)
        was expecting correct count. Thus mixing WORDOUT's and OUT's
        on the same channel screwed up. IOSER p. 14
\UR#24\ JRL 5/3/78. Export (TOPS-10) interrupt handler for floating
        underflow was bombing because PDL was not being restored
        properly. TRIGS p 4.

\UR#25\ JRL 6/20/78. New feature. Allow # followed by octal digits
        to be used as filenames and extensions. Added so that SAIL
        programs can read UFD's where PPN if interpreted in sixbit
        contained a period, [, ] or comma. With this SAIL can read
        and write any filenames (not just those which are legal
        identifiers). IOSER p 7

\UR#26\ JRL 8/2/78. ARRAY subscripting code (SUBSCR) was throwing
        away the STRING TBITS bit for STRING ITEMVAR ARRAYS
        Fix was to explicitly check for itemvar each place
        string was being checked. (2 removals of clearing STRING
        bit, 3 additions of check of itemvar). 
        ARRAY page 15.
\UR#27\ JRL (8/2/78) - ARMAK screws up if array to be allocated
        is supposed to have more than 256K words. Inserted
        error check to print ARRAY NO ROOM message.
        ARRSER/2

\UR#28\ JRL(8/4/78) - Allow string expansion to use max of STINCR
        and length of new string for size of new area. GOGOL
        EXPSTR+2 p 49

        Note this fixes one problem but causes another. Since
        the compactifier always uses the string areas in the order
        they were allocated (and only releases the most recently
        allocated one) we can waste a lot of space. Solution
        would be to release a string block
        if no strings were stored in it.

\UR#29\ JRL (8-7-78) - Release string blocks containing no strings
        during the garbage collector. Can't release last string block
        (since it is where new string will be placed). New field
        (.TOPBYTE) in the string block headers to allow us to know
        if block contains strings. GOGOL (5 changes)
        
\UR#30\ JRL (8-7-78) \UR#29\ TICKLED BUG THAT A CONSTANT (4) HAD
        BEEN USED INSTEAD OF THE SYMBOLIC NAME .HDRSIZ  
        GOGOL/21



\UR#31\ JRL (8-9-78) install NIH switch. HEAD, SCISS.SAI, SAIL 
        Make /26A default for NIH.

\UR#32\ JRL (8-10-78) make FORTRAN-10 and KL10 opcodes default
        Make /35A switch default for UOR

\UR#33\ JRL (8-15-78) TRAGO GOTO-SOLVER WASN'T COUNTING BLOCK
        LEVELS CORRECTLY. ANY BLOCK WITH DECLARATIONS HAS AN
        LVI ENTRY AND MUST BE COUNTED. CHECK IS NOW ONLY TO
        SEE IF ANY DECLARATIONS (FORMERLY ONLY CHECKED FOR
        SETS AND ARRAYS) . STATS/15

\UR#34\ JRL (9-14-78) bug in fix \UR#33\ don't put out bexit
        is simple procedure. fix check simpsw at DOBLK in
        TRAGO code. STATS/15

\UR#35\ JRL (10-31-78) fixed REF!ITEM to allow arrays.
	HEL near RF0 and LEPRUN at REFITM

\UR#36\ JRL (12-22-78) CONELM SCREWED UP SBITS OF PTR RESULT
        NEED TO TURN OFF EXTRN INTRN BITS. SHOWED UP WHEN
        ASSIGNING FUNCTION VALUE TO CONTEXT ELEMENT WHICH
        REFERED TO EXTERNAL VARIABLE

             CON:EXTRN := FUNCTION
  
        LEAP PAGE 27. CONELM + 13
\UR#37\ JRL (1-12-79) BUG FIX \UR#35\ BUGGY. TREATED ALL
       REF!ITEMS AS ARRAY REFITEMS. LEPRUN  NEAR REFITM PAGE 46.