1 
 2 09/23/87  analyze_multics
 3 Known errors in the current release of analyze_multics.
 4 #         Associated TR's
 5 Description
 6 
 7 70
 8 Display on units which occur in stack.
 9 
10 69
11 When the "-for" control argument is given with requests such as "stack"
12 for example, integers greater than seven (7) are not allowed.
13 
14 68  phx20553
15 The structure name "dte" when used as "display -as dte" uses an
16 incorrect structure to format the output.  Instead of using ioi_dte as
17 the structure_names info segment states, it uses disk_table_entry for
18 formatting.
19 
20 67  phx20262
21 A malformed ioa_ control string is displayed when the select_dump
22 request is used and the referenced dumps do not exist.
23 
24 66  phx20247
25 azm why request aborts if a segment number is found to be out of the
26 range of segment numbers.
27 
28 65  phx20120
29 The crash message is not printed any longer since BCE was implemented.
30 
31 64  phx19721
32 azm display request will not dump a segment which is not in the fdump,
33 even if an online copy is available.
34 
35 63  phx19335
36 User's process is destroyed if too many dumps are referenced within the
37 same azm invokation.  The process dir get too filled up.
38 
39 62  phx19331
40 The azm stack request aborts when it can't find the stack.
41 
42 61  phx19329
43 The azm display request does not dump segments that are not in the
44 fdump without specifying -instruction as a control arg.  It should by
45 default dump a "not dumped" segment based on the online copy after
46 stating it cannot be found in the dump.
47 
48 60  phx19327
49 mc -prds sys -prs 2 thinks the "2" is a virtual address.
50 
51 59  phx18934
52 The azm replace request does not work when attempting to replace
53 hardcore segments.
54 
55 55  phx18803
56 azm_why_ aborts to multics command level via amu_error_ instead of just
57 aborting the request.
58 
59 53  phx18785
60 The "stack -ag" request mis-aligns the arguments of a stack frame when
61 printed.
62 
63 52  phx18684
64 The search request miscaluates the end of a segment (example is
65 dn355_mailbox, see TR for details).  display calculates it correctly
66 tho.
67 
68 45
69 Should not attempt to print machine conditions that are obviously
70 incorrect; a good test of this might be that all of the "pointer
71 registers" in the machine conditions do not have ITS tags where they
72 should.  Perhaps a -force control argument to get around this would
73 help.
74 
75 This comes up in cases such as doing "mc pr6" to look at machine
76 conditions in the current stack frame, followed by another "mc pr6"--
77 since the first one set "pr6" to something different, the second one
78 should fail, and, indeed, it does, but by printing trash, rather than
79 any useful diagnostic.
80 
81 29  phx17437
82 In the search request, the "Segment is not in the fdump" error message
83 has garbage in it (nulls) and also incorrectly references the segment
84 name in the msg.
85 
86 Also, a "srh foo 555555555555" where foo is not a segment in the dump,
87 produces an error message "Invalid argument" instead of saying "Segment
88 foo not found in fdump" or somesuch.