Title: | DECmcc user notes file. Does not replace IPMT. |
Notice: | Use IPMT for problems. Newsletter location in note 6187 |
Moderator: | TAEC::BEROUD |
Created: | Mon Aug 21 1989 |
Last Modified: | Wed Jun 04 1997 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 6497 |
Total number of notes: | 27359 |
It seems MCC T1.2.7 broke mcc_ilv_put for the MCC_K_DT_FULL_ENTITY datatype when the reason code is not MCC_K_REASN_SUCC_OP. I use to encode a 'non-available' full entity datatype with reason_code = MCC_K_REASN_ATTR_NOT_AVAIL p_mcc_desc->mcc_a_pointer = MCC_K_NULL_PTR p_mcc_desc->mcc_w_curlen = 0 now it result in wierd results from both FCL and IMPM, although the ILV dump is still the same as in T1.2.4 (which worked fine). in the example below, Forward Queue is a full entity name datatype: ^^^^^^^^^^^^^ Here what T1.2.4 had Type = Full QueueType Version = 1 Volatile = False Temporary = False Consistency = Relaxed ElementDefName = -- Attribute Not Available SelectionQual = -- Attribute Not Available Forward Queue = -- Attribute Not Available MaxNumber Registrations = 0 MaxNumber Elements = -- Attribute Not Available MaxSize Element = 0 Bytes Locked Operations = { } Here what T1.2.7 gives Type = Full QueueType Version = 1 Volatile = False Temporary = False Consistency = Relaxed ElementDefName = -- Attribute Not Available SelectionQual = -- Attribute Not Available, Note here the comma after the last attribute and note that all other attributes disappeared (not displayed) although they are present in the ILV buffer. Below is a partial out_p dump where the IDCodes 44 corresponds to SelectionQual, 46 corresponds to Forward Queue, and 64 corresponds MaxNumber Registrations [ 44 ] ( [ 1 ] 1a [ 2 ] [ 3 ] 03 ) [ 46 ] ( [ 1 ] 37 -- 7 [ 3 ] 03 [ 4 ] ( ) ) [ 64 ] ( [ 1 ] 23 -- # [ 2 ] 00 [ 3 ] 00 ) Note that if I remove the 'Forward Queue' attribute (IDCode 46) from the buffer, everything gets displayed correctly with T1.2.7 Please treat this as a high priority request in checking it since T1.2.7 is the MCC kit that our External Field customers started receiving. Luckily the AM has gone out yet, so i need to put a fix or a workaround. Thanks, Phil
T.R | Title | User | Personal Name | Date | Lines |
---|---|---|---|---|---|
2862.1 | the problem is in mcc_ilv_get, for NULL full entities only. | TOOK::KOHLS | Ruth Kohls | Thu Apr 30 1992 14:57 | 14 |
>It seems MCC T1.2.7 broke mcc_ilv_put for the MCC_K_DT_FULL_ENTITY datatype ... The problem is in mcc_ilv_>get<, for completely NULL full entities only. (example: something declared: MCC_A_AES full_entity = NULL; and then mcc_aes_create(...) never used on it.) The problem will occur whether or not they are in an attrib_list, so whether or not there is a reason code involved. A temporary workaround is to not ILV-encode completely null entities. The problem will be fixed soon, though, so the workaround is temporary. Ruth K. |