[Search for users]
[Overall Top Noters]
[List of all Conferences]
[Download this site]
| Title: | DEC Ada |
| Notice: | Ada is no longer a trademark of the US Government |
| Moderator: | KMOOSE::CMCCUTCHEON |
|
| Created: | Mon Jan 27 1986 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 3874 |
| Total number of notes: | 16668 |
3849.0. "DEBUGGER, EXAMINE does not show whole record" by TAV02::ZVI_P (Here we are) Wed Mar 05 1997 08:27
Hi
A customer problem:
The debugger EXAMINE command does not show all the fields of a record.
The customer supplied a reproducer, we reproduced it on OVMS ALPHA 6.2
with ALPDEBU05_070 (debugger patch) and ADA 3.3-7 .
problem will be described by a log of a session.
ORTAL_PALGY> run digital_check
OpenVMS Alpha DEBUG Version V6.2-103R
%DEBUG-I-INITIAL, language is ADA, module set to DIGITAL_CHECK
%DEBUG-I-NOTATMAIN, type GO to get to start of main program
DBG> Go
break at routine DIGITAL_CHECK
3: procedure digital_check is
DBG> set br %lin 25
DBG> typ 25
module DIGITAL_CHECK
25: For Counter in 1..20 loop
DBG> g
break at DIGITAL_CHECK.%LINE 25
25: For Counter in 1..20 loop
DBG> Step
stepped to DIGITAL_CHECK.BLOCK$25.%LINE 26
26: Data_Array(Counter) := "DROR ";
DBG> Step
stepped to DIGITAL_CHECK.BLOCK$25.%LINE 27
27: Digital_Check.Unchecked_Insert(list=> Digital_List,
DBG> Step
stepped to DIGITAL_CHECK.BLOCK$25.%LINE 31
31: end loop;
DBG> ex digital_list
DIGITAL_CHECK.DIGITAL_LIST
SIZE: 20
LENGTH_ALLOCATED: 1
LENGTH_USED: 1
FREE_LIST_FIRST_IDX: 0
FREE_LIST_LAST_IDX: 0
USED_LIST_FIRST_IDX: 1
USED_LIST_LAST_IDX: 1
PREV_HANDLE
(1): 0 <<<<*********** 19 more elements missing.
******>>>>
***>>>>>>> Also missing are the components:
***>>>>>>>
***>>>>>>> Next_Handle, Key_Warehouse and Data_Warehouse
DBG> sho symb/typ PREV_HANDLE
record component DIGITAL_CHECK.LIST_TYPE.PREV_HANDLE
noncontiguous array descriptor type, 1 dimension, bounds: [1:1], size:
2 byt
es
cell type: subrange type, size: 2 bytes, range: 0..20
parent type: atomic type, word integer, size: 2 bytes
DBG>
*******>>>> The debugger wrongly parses the data type ****
The definition of "Digital_List" is:
In File: DIGITAL_CHECK.ADC
package Digital_check is new static_list(Integer_Range,
Data_Type,
max_list_length,
minimum_free_list_size,
complex_packing);
subType Digital_List_Type is Digital_Check.List_Type;
Integer_Range_Data : Integer_Range;
Digital_list : Digital_List_Type;
In File: STATIC_LIST_.ADC
type List_Type (Size : Length_Type := Length_Type'last) is record
Length_Allocated : Length_Type := 0; -- position of the "farest"
-- element in buffer
Length_Used : Length_Type := 0; -- number of elements currently
-- in buffer
Free_List_First_Idx : Handle_Type := NULL_HANDLE;
Free_List_Last_Idx : Handle_Type := NULL_HANDLE;
Used_List_First_Idx : Handle_Type := NULL_HANDLE;
Used_List_Last_Idx : Handle_Type := NULL_HANDLE;
Prev_Handle : Pointers_Type(1 .. Size);
Next_Handle : Pointers_Type(1 .. Size);
Key_Warehouse : Keys_Warehouse_Type(1 .. Size);
Data_Warehouse : Data_Warehouse_Type(1 .. Size);
end record;
On OVMS VAX 6.2 with VAXDEBU05_070 and ADA 3.3-8 this works fine:
...
DBG> ex Digital_List
DIGITAL_CHECK.DIGITAL_LIST
SIZE: 20
LENGTH_ALLOCATED: 1
LENGTH_USED: 1
FREE_LIST_FIRST_IDX: 0
FREE_LIST_LAST_IDX: 0
USED_LIST_FIRST_IDX: 1
USED_LIST_LAST_IDX: 1
PREV_HANDLE
(1): 0
(2): 31548
(3): 0
(4): 11080
(5): 32748
(6): -28925
(7): 0
(8): 11080
(9): 32748
(10): 29926
(11): 0
(12): 11080
(13): 32748
(14): -29601
(15): 0
(16): 11080
(17): 32748
(18): 29275
(19): 0
(20): 11080
NEXT_HANDLE
(1): 0
(2): -30234
(3): 0
(4): 11080
(5): 32748
(6): 29112
(7): 0
(8): 11080
(9): 32748
(10): -30794
(11): 0
(12): 11080
(13): 32748
(14): 28996
(15): 0
(16): 11080
(17): 32748
(18): -31561
(19): 0
(20): 11080
KEY_WAREHOUSE
(1): 1
(2): 726138880
(3): -2092269588
(4): 726138880
(5): 1869185004
(6): 726138880
(7): -2115403796
(8): 726138880
(9): 1858109420
(10): 726138880
(11): 2145484780
(12): 726138880
(13): 1848999916
(14): 726138880
(15): 2126479340
(16): 726138880
(17): 1842380780
(18): 726138880
(19): -1848541204
(20): 726138880
DATA_WAREHOUSE
(1)(1..10): "DROR "
(2)(1..10): "........k"
(3)(1..10): "..H+�.Z..."
(4)(1..10): "H+�.Qk..H+"
(5)(1..10): "�.+...H+�."
(6)(1..10): "Wj..H+�.x."
(7)(1..10): "..H+�..i.."
(8)(1..10): "H+�.}...H+"
(9)(1..10): "�.�h..H+�."
(10)(1..10): "...H+�.g"
(11)(1..10): "..H+�.�..."
(12)(1..10): "H+�.bg..H+"
(13)(1..10): "�.�...H+�."
(14)(1..10): "�f..H+�.."
(15)(1..10): "..H+�.}.."
(16)(1..10): "H+�.�...H+"
(17)(1..10): "�.<:..�.."
(18)(1..10): ".........."
(19)(1..10): "..... �.�."
(20)(1..10): "..�..�..&�"
The sources are at:
$ dir ppantr::$2$DKB200:[IPMT.ADA_DBG2]
Directory PPANTR::$2$DKB200:[IPMT.ADA_DBG2]
ARRAY_STORAGE_MANAGEMENT.ADC;2
5/7 3-MAR-1997 10:57:46.00
ARRAY_STORAGE_MANAGEMENT_.ADC;2
3/7 3-MAR-1997 10:57:40.00
DIGITAL_CHECK.ADC;8
4/7 3-MAR-1997 11:10:49.00
HASH_TYPES_.ADC;1 2/7 3-MAR-1997 10:51:21.00
INTEGER_LNM.ADC;1 1/7 3-MAR-1997 10:49:08.00
INTEGER_LNM_.ADC;1 2/7 3-MAR-1997 10:47:24.00
IS_BLANK.ADC;1 1/7 3-MAR-1997 10:54:18.00
LIST_HANDLE_TYPES_.ADC;1
2/7 2-MAR-1997 16:39:56.00
LNM.ADC;1 6/7 3-MAR-1997 10:48:54.00
LNM_.ADC;1 3/7 3-MAR-1997 10:48:49.00
PACK_TYPES.ADC;5 9/14 3-MAR-1997 10:58:14.00
PACK_TYPES_.ADC;3 18/21 3-MAR-1997 10:58:06.00
REGULAR_HASHING.ADC;1
5/7 3-MAR-1997 10:46:05.00
REGULAR_HASHING_.ADC;1
2/7 3-MAR-1997 10:45:58.00
SEMAPHORE.ADC;1 2/7 3-MAR-1997 11:01:31.00
SEMAPHORE_.ADC;1 2/7 3-MAR-1997 11:01:26.00
STANDARD_PRINT.ADC;2
18/21 3-MAR-1997 10:55:06.00
STANDARD_PRINT_.ADC;2
10/14 3-MAR-1997 10:54:58.00
STATIC_LIST.ADC;1 95/98 2-MAR-1997 16:49:12.00
STATIC_LIST_.ADC;1 63/63 2-MAR-1997 16:40:08.00
STORAGE_HASHED_LIST.ADC;1
45/49 3-MAR-1997 10:51:57.00
STORAGE_HASHED_LIST_.ADC;1
23/28 3-MAR-1997 10:51:51.00
STORAGE_TRACE.ADC;1
13/14 3-MAR-1997 11:03:13.00
STORAGE_TRACE_.ADC;3
4/7 3-MAR-1997 11:03:05.00
STORAGE_TRACE__ADD.ADC;1
3/7 3-MAR-1997 11:08:11.00
STORAGE_TRACE__CONVERT_NAME.ADC;1
2/7 3-MAR-1997 11:06:12.00
STORAGE_TRACE__POINT.ADC;1
2/7 3-MAR-1997 11:07:38.00
STORAGE_TRACE__PRINT_ADDRS.ADC;1
7/7 3-MAR-1997 11:09:04.00
STORAGE_TRACE__PRINT_TRACE_LINE.ADC;1
6/7 3-MAR-1997 11:07:03.00
STORAGE_TRACE__REMOVE.ADC;1
3/7 3-MAR-1997 11:08:42.00
TIME_DEFAULTS_.ADC;1
3/7 2-MAR-1997 16:43:22.00
TIME_SERVICES.ADC;2
42/42 3-MAR-1997 10:54:26.00
TIME_SERVICES_.ADC;2
14/14 3-MAR-1997 10:53:24.00
Total of 33 files, 420/532 blocks.
$
Note: files type is ADC .
After compilation link by:
$ acs lin/deb DIGITAL_CHECK
DIGITAL_CHECK is the image to be created.
Thanks
Zvi
| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 3849.1 | Restriction; text from V3.3 release notes | KMOOSE::CMCCUTCHEON | Charlie McCutcheon | Wed Mar 05 1997 11:06 | 52 |
| Hi, I'm afraid this is a debugging restriction. From the V3.3
release notes:
5.6 Debugging Restrictions
The following restrictions apply when you debug DEC Ada
programs:
...
o A problem occurs when a record's discriminant is used
for the size of a record component. In the following
example:
type VAR_STRING (LEN : INTEGER) is record
VALUE : STRING (1..LEN);
end record;
V : VAR_STRING(10) := (10, "1234567890");
Given the command EXAMINE V.VALUE, the debugger
displays one character for V.VALUE.
The workaround is to give the length of the character
string in the debugger command line as in the following
example:
EXAMINE /asci=10 V.VALUE
...
o Record components following variable-sized components
cannot be debugged. Digital suggests that if possible
you reorder the components in your program for
debugging purposes.
In your user's program:
type List_Type (Size : Length_Type := Length_Type'last) is record
...
Prev_Handle : Pointers_Type(1 .. Size);
...
This is the same documented problem, the user is using the record discrimimant
as the size of one of the record components, so the size of that component
is not known by Debug, and the offset to any following record component
is not known either.
Unfortunately, I don't see this restriction ever changing. This requires
work from the GEM code generator group, and they are busy working on
nifty baselevels that DEC Ada may never see, so new support for this issue
is highly unlikely.
Charlie
|