Summary: | <net-analyzer/net-snmp-5.4 fails to build | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Kenton Groombridge <rustyvega> |
Component: | New packages | Assignee: | Gentoo Netmon project <netmon> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | amari, anf, aross, corsair, hejian.he, ian, ierdnah, jaco, just_john, old-stone, pawel.mandes, zubkov318 |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | 163094 | ||
Bug Blocks: | |||
Attachments: | net-snmp-defined.patch |
Description
Kenton Groombridge
2006-11-24 22:09:46 UTC
I had the same problem, and oddly running it through cpp with the right include dirs made it compile. cpp -I /var/tmp/portage/net-analyzer/net-snmp-5.3.1-r1/work/net-snmp-5.3.1/agent/mibgroup/ucd-snmp -I /var/tmp/portage/net-analyzer/net-snmp-5.3.1-r1/work/net-snmp-5.3.1/agent/mibgroup /var/tmp/portage/net-analyzer/net-snmp-5.3.1-r1/work/net-snmp-5.3.1/agent/mibgroup/ucd-snmp/disk.c > disk.c Seems to be fix in version 5.4 *** Bug 160250 has been marked as a duplicate of this bug. *** I can confirm this error with net-snmp-5.2.1.2-r1 (latest stable version). Using net-snmp-5.4 from ~x86 worked for me. *** Bug 160899 has been marked as a duplicate of this bug. *** *** Bug 161246 has been marked as a duplicate of this bug. *** *** Bug 161314 has been marked as a duplicate of this bug. *** Reopen; either stabilize something working (5.4) or fix this please. Created attachment 106348 [details, diff]
net-snmp-defined.patch
This patch makes it compile for me, should apply cleanly with patch -p1, guess this should be sufficient to allow maintainer to create a -r2 and get this resolved, thanks.
Same issue. Added mysqelf to CC. *** Bug 161919 has been marked as a duplicate of this bug. *** Same issue with 5.2.1.2-r1. Further up in my log, specifically in the "checking for platform-specific source..." secion, I see: checking for library containing getmntent... none required Which might mean that the value is being ignored rather than set to 0. Would it be better to correct that in case some other cases actually behave differently when it is set to 0? I believe this patch would break those. Sorry, don't know enough to suggest a different patch. *** Bug 163168 has been marked as a duplicate of this bug. *** 5.4 stabilization handled in Bug 163094, closing again. *** Bug 163719 has been marked as a duplicate of this bug. *** *** Bug 164380 has been marked as a duplicate of this bug. *** It would still be nice to fix this bug in 5.2.1.2-r1 at least. I traced through some more of the code built a similar patch, which wraps all occurrences of HAVE_GETMNTENT, HAVE_STATFS, HAVE_STATVFS, and HAVE_SETMNTENT with defined(). It looks like these variables are supposed to be empty (or simply #DEFINE'd), so I'm guessing the error shows up now because the newer GCCs substitute and then process the ||, whereas the older ones probably didn't (or something similar)... |