David Lamparter
David Lamparter
Throwing userspace programs with atomic types at pahole currently results in either odd results (struct fiels with zero length) or straight up segfaults (dereferencing a NULL tag). Add DW_TAG_atomic_type analogous...
This merges the most active forks I could find & adds standard autoconf/automake build.
As noted in commit text: This allows invoking an external wrapper script with config on stdin/out and frr.conf on stdout/in for saving/loading respectively. Example use cases would be doing a...
Unfortunately this isn't really done as-is since the ioctl value does not uniquely identify this ioctl. Some way to check the socket AF is needed to disambiguate :( --- If...
Related to #3615 - there's a bunch of socket family specific ioctls: ``` $ pcregrep -r 'SIOCPROTOPRIVATE\s*\+\s*1\W' /usr/include/linux* /usr/include/linux/ax25.h:#define SIOCAX25ADDUID (SIOCPROTOPRIVATE+1) /usr/include/linux/rose.h:#define SIOCRSSCAUSE (SIOCPROTOPRIVATE+1) /usr/include/linux/kcm.h:#define SIOCKCMUNATTACH (SIOCPROTOPRIVATE + 1) /usr/include/linux/mroute6.h:#define...
xref.h says: ``` /* this macro is invoked once for each standalone DSO through * FRR_MODULE_SETUP \ * }-> FRR_COREMOD_SETUP -> XREF_SETUP * FRR_DAEMON_INFO / */ #define XREF_SETUP() … ```...
This can be used to get less cryptic error/warnings from GCC when dealing with something typesafe container related. --- as requested by @mjstapp :wink:
changes to the interface type weren't correctly handled in updating the connected routes - changing the interface type wasn't reoriginating connected routes to begin with, meaning stale routes could be...
The gcc plugin was already warning about printfrr'ing `time_t`, but `struct timeval` uses `suseconds_t`. - fix the one place where this matters in lib/NB code - add `__suseconds_t`/`__suseconds64_t` to things...