Bug 877955
| Summary: | [abrt] cscope-15.8-2.fc16: __GI_raise: Process /usr/bin/cscope was killed by signal 6 (SIGABRT) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Burhan Ali <burhan.ali> | ||||||
| Component: | cscope | Assignee: | Neil Horman <nhorman> | ||||||
| Status: | CLOSED NOTABUG | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||||
| Severity: | unspecified | Docs Contact: | |||||||
| Priority: | unspecified | ||||||||
| Version: | 16 | CC: | nhorman | ||||||
| Target Milestone: | --- | ||||||||
| Target Release: | --- | ||||||||
| Hardware: | x86_64 | ||||||||
| OS: | Unspecified | ||||||||
| Whiteboard: | abrt_hash:fc4530bb6a7850329cfe91add0816660581ab6a6 | ||||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||||
| Doc Text: | Story Points: | --- | |||||||
| Clone Of: | Environment: | ||||||||
| Last Closed: | 2012-11-19 16:31:34 UTC | Type: | --- | ||||||
| Regression: | --- | Mount Type: | --- | ||||||
| Documentation: | --- | CRM: | |||||||
| Verified Versions: | Category: | --- | |||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||
| Embargoed: | |||||||||
| Attachments: |
|
||||||||
|
Description
Burhan Ali
2012-11-19 10:26:00 UTC
Created attachment 647665 [details]
File: environ
Created attachment 647666 [details]
File: backtrace
The inverse library that failed the assertion on this failure appears to have failed because it contains indexing structures of a different size than the local system thinks they should be - Was the inverted database built on a different system? I don't fully understand the question, but the answer is likely to be Yes. The working directory in question is on a development server (32-bit CentOS) where the cscope files were built. This directory is mounted onto my desktop machine (64-bit Fedora) over NFS where I am editing some of the files. (In reply to comment #4) > I am editing some of the files. Actually this isn't true. Just being in that directory and running vim is enough to trigger the crash. I don't need to be editing any of the files referenced by cscope. [burhan@orion code]$ vim Error detected while processing /etc/vimrc: line 39: cs_read_prompt EOF: Success E609: Cscope error: cscope: invlib.c:570: invopen: Assertion `invcntl-aram.sizeblk == sizeof(t_logicalblk)' failed. Press ENTER or type command to continue The section of /etc/vimrc mentioned in the error message is below: 32 if has("cscope") && filereadable("/usr/bin/cscope") 33 set csprg=/usr/bin/cscope 34 set csto=0 35 set cst 36 set nocsverb 37 " add any database in current directory 38 if filereadable("cscope.out") 39 cs add cscope.out 40 " else add database pointed to by environment 41 elseif $CSCOPE_DB != "" 42 cs add $CSCOPE_DB 43 endif 44 set csverb 45 endif Thats ok, you answered the question appropriately. The long and the short of it is that cscope, being a very old piece of code, creates database file, in which its index metadata is sized relative to the architecture that its running on, so when you create a database on a 32 bit system and attempt to read it on a 64 bit system, it fails with the above assertion. We should probably make it exit gracefully (or better still make the data structure size agnostic to system arch), which I would love to get to upstream, but for now this is working as designed. The proper course of action is to only use a given database on simmilar arches. |