| Anonymous | Login | 10-23-2019 00:17 UTC |
| Main | My View | View Issues | Change Log | Docs |
| Viewing Issue Simple Details [ Jump to Notes ] | [ View Advanced ] [ Issue History ] [ Print ] | ||||||||
| ID | Category | Severity | Reproducibility | Date Submitted | Last Update | ||||
| 0004313 | [Squeak] Compiler | major | always | 07-22-06 23:18 | 02-06-11 23:48 | ||||
| Reporter | kwl | View Status | public | ||||||
| Assigned To | nicolas cellier | ||||||||
| Priority | normal | Resolution | fixed | ||||||
| Status | closed | Product Version | 3.9 | ||||||
| Summary | 0004313: Compiler+Decompiler together have a serious 'bad case' bug | ||||||||
| Description |
In Decompiler there is a situation which raises an error, located in #send:super:numargs. When this erros occurs the emergency console comes up because the debugger fails because the decompiler fails because something else fails (dunno what). Unfortunately, emergency situations are almost impossible to debug, so I can't attach any further information :( The situation can be reproduced by evaluating the following innocent looking snippet, took quite some time to find it </phew> : | x y | x := y := 0. {'a'. 'b'. 'c'} do: [:c| (c string halt; endsWith: 'd') ifTrue: [x:= 1 + x] ifFalse: [y:= 1 + y]] Checked in 3.8 and 3.9, same result. |
||||||||
| Additional Information | |||||||||
| Attached Files |
 decompileCascade-bf.1.cs.gz [^] (864 bytes) 07-23-06 16:16 |
||||||||
|
|
|||||||||
 Relationships |
|
|
Notes |
|
|
(0006182 - 179 - 336 - 336 - 336 - 336 - 336) bert 07-23-06 16:08 |
Simpler test case: test: a (a foo; bar) ifTrue: [1] ifFalse: [2] If the receiver of an ifTrue:ifFalse: is a cascade, the decompiler mistakes it for a case statement. |
|
(0006183 - 22 - 22 - 22 - 22 - 22 - 22) bert 07-23-06 16:16 |
Attached proposed fix. |
|
(0006185 - 134 - 134 - 134 - 134 - 134 - 134) kwl 07-23-06 23:24 |
The proposed fix works for the two examples. Is there a routine for checking the modified Decompiler against all source coded methods? |
|
(0013867 - 120 - 132 - 132 - 132 - 132 - 132) nicolas cellier 09-07-10 19:53 |
In 4.2 trunk, Decompiler still fails with Klaus example. Consequences are not catastrophic though. Bert example works. |
|
(0013869 - 83 - 83 - 83 - 83 - 83 - 83) nicolas cellier 09-07-10 21:49 |
Oops; my mistake, both issues are fixed in latest 4.2 trunk and probably in 4.1 too |
|
Issue History |
|||
| Date Modified | Username | Field | Change |
| 07-22-06 23:18 | kwl | New Issue | |
| 07-23-06 16:08 | bert | Note Added: 0006182 | |
| 07-23-06 16:16 | bert | File Added: decompileCascade-bf.1.cs.gz | |
| 07-23-06 16:16 | bert | Note Added: 0006183 | |
| 07-23-06 23:24 | kwl | Note Added: 0006185 | |
| 09-07-10 19:53 | nicolas cellier | Note Added: 0013867 | |
| 09-07-10 21:49 | nicolas cellier | Status | new => resolved |
| 09-07-10 21:49 | nicolas cellier | Fixed in Version | => trunk |
| 09-07-10 21:49 | nicolas cellier | Resolution | open => fixed |
| 09-07-10 21:49 | nicolas cellier | Assigned To | => nicolas cellier |
| 09-07-10 21:49 | nicolas cellier | Note Added: 0013869 | |
| 02-06-11 23:48 | leves | Status | resolved => closed |
| Mantis 1.0.8[^]
Copyright © 2000 - 2007 Mantis Group
57 total queries executed. 41 unique queries executed. |  |
