1
2 09/21/87 gcos
3 Known errors in the current release of gcos.
4 # Associated TR's
5 Description
6
7 1090 phx18132
8 The DRUN facility of gtss is making some invalid assumptions about what
9 to return when a DRL is executed from a DRUN. This can cause RUNOFF to
10 fail if executed in a DRUN.
11
12 1089 phx17058
13 Physical record processing in $UTILITY will usually result in an abort.
14 This is because the gcos simulator does not handle multiple dcw lists
15 correctly for tape IO. Tape IO with lists of IOTPs will fail.
16
17 1088 phx16384
18 The gcos simulator may not return the correct data in the A register on
19 sucessful completion of a MME GECALL.
20
21 1087 phx08471
22 Simulation of MME GECALL and MME GERSTR do not give valid results if a
23 non-zero load origin is given.
24
25 1086 phx07434
26 If a file allocated by gtss grows to a msf, its unique id in the mcfc
27 data base will change allowing new allocations to bypass the mcfc
28 mechanism.
29
30 1085 phx04086
31 The gcos_tss multiple caller file control mcfc data base sometimes has
32 entries retained after the caller exits gtss. Some of these entries are
33 for directories which should never appear. There should be no entries
34 for a caller who has exited gtss.
35
36 1084 phx03121
37 The gcos batch simulator will not process H* files from tape. These
38 files must be true disk files to be used properly.
39
40 1073 phx15715
41 The gtss mcfc database is unprotected. The segments GTSS.MCFC.* have
42 rew access to *.*.*. This is necessary for gtss to provide concurrency
43 control for gtss files. These mcfc files are not accessable to gtss
44 users normally
45
46 1066 phx15272
47 If a gcos batch job quits abnormally and leaves the sysprint file
48 unclosed, gcos_sysprint cannot print it. Instead, gcos_sysprint will go
49 into an infinite loop reporting that it has reached the end of
50 information.
51
52 1062 phx15034 phx15266
53 If GTSS is invoked while using window_call or video mode it puts out
54 the error message that it can't get the terminal attributes and quits.
55
56 1060 phx19399
57 A crun in gtss will terminate output cout after about 140 lines as if
58 this was the proper thing to do. Remaining output is thrown away
59 somewhere.
60
61 1044 phx02637
62 Run units H* ** created by the batch simulator via MME GESAVEcannot
63 be used by gtss. The batch simulator H*'s are not native GCOS mode but
64 gtss requires native GCOS H*'s.
65
66 1039 phx14478
67 Using GFMS may fail due to no more room in the known seg table KST.
68
69 1022 phx12722
70 Objects to option fields on certain control cards if they are over 6
71 characters long. Since the real GCOS accepts longer fields, so should
72 we.
73
74 1021 phx12303 phx12802 phx17231
75 When writting to tape, GCU fails to flush out the last buffer if it is
76 not full. This results in leaving off data from the end of a tape file.
77 Also, it is doing a less-than-optimal job of translating data to GCOS
78 BCD format.
79
80 1014 phx03694
81 YFORTRAN doesn't function at all. Just says "please retype last
82 command".
83
84 1013 phx03920
85 gcos_user$print only accepts 1 argument.
86
87 1009 phx07777 phx07604 phx01933
88 If a tape drive is assigned that provides both density 1600 and density
89 6250 the gcos "UTILITY" program institutes a label checking procedure
90 that is assured to fail. This is a result of its checking density
91 dynamically and not processing the hardware i/o status codes properly.
92 If a drive with density 1600 and less is assigned failure will not
93 occur. It is not possible to use Multics "resource control" as the
94 simulator implements the functionality required by gcos at the MME
95 level.