Patch Set 9: Code-Review+2
--
To view, visit https://gerrit.osmocom.org/4593
To unsubscribe, visit https://gerrit.osmocom.org/settings
Gerrit-MessageType: comment
Gerrit-Change-Id: Ic77866ce65acf524b768882c751a4f9c0635740b
Gerrit-PatchSet: 9
Gerrit-Project: libosmocore
Gerrit-Branch: master
Ger
Harald Welte has submitted this change and it was merged.
Change subject: Enable GnuTLS fallback
..
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id()
Patch Set 9:
OS#2610 is created to track optimization suggestion.
--
To view, visit https://gerrit.osmocom.org/4593
To unsubscribe, visit https://gerrit.osmocom.org/settings
Gerrit-MessageType: comment
Gerrit-Change-Id: Ic77866ce65acf524b768882c751a4f9c0635740b
Gerrit-PatchSet: 9
Gerrit-Projec
Hello Harald Welte, Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#9).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would alw
Patch Set 8: Code-Review+2
(1 comment)
I'm happy to merge it, but my original review comments about using a static
variable to cache whether the syscall works has not been implemented. Please
follow up with that in a later patch (or create a ticket as a reminder for this
optimization)
https:
Patch Set 8:
That's odd - why there're 2 rebuilds triggered by single gerrit push?
--
To view, visit https://gerrit.osmocom.org/4593
To unsubscribe, visit https://gerrit.osmocom.org/settings
Gerrit-MessageType: comment
Gerrit-Change-Id: Ic77866ce65acf524b768882c751a4f9c0635740b
Gerrit-PatchSet
Hello Harald Welte, Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#8).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would alw
Hello Harald Welte, Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#7).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would alw
Patch Set 6: Code-Review-1
* getrandom() and SYS_getrandom only used if detected at compile time
* if either of those returns an error at runtime, fall back to gnutls (if
present/selected at compile time)
--
To view, visit https://gerrit.osmocom.org/4593
To unsubscribe, visit https://gerrit.os
Hello Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#6).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would always return fai
Hello Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#5).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would always return fai
Patch Set 3:
I've updated comments and commit message. My confusion comes from following:
currently we check for glibc version and syscall availability at compile time -
and use either of those if available. The GnuTLS fallback is added in the same
way: if compile-time check for getrandom (bot
Hello Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#3).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would always return fai
Patch Set 2:
> I am arguing for a "trivial fallback" but in a way that a single binary will
> determine at runtime if getrandom() is available, or if not, fall back to
> gnutls.
Ah, get it now - misunderstood your comment altogether. Will send updated patch
shortly.
--
To view, visit https:
Patch Set 2:
> > I would argue to activate the --enable-gnutls by default, if it
> is present on the system.
>
> > Or maybe rather, fail configure.ac if gnutls is not found AND no
> --disable-gnutls was present.
>
> What should be the default for .deb packages?
a run-time fallback to gnu
Patch Set 2:
> Could you elaborate why that's more efficient?
Isn't it the cost of a syscall (very expensive) vs. the cost of a single
conditional if statement? Or am I missing something?
> AFAIK with the way we use getrandom, it can only fail permanently
yes, exactly. This means that on
Patch Set 2:
> I would argue to activate the --enable-gnutls by default, if it is present on
> the system.
> Or maybe rather, fail configure.ac if gnutls is not found AND no
> --disable-gnutls was present.
What should be the default for .deb packages?
--
To view, visit https://gerrit.osmoc
Patch Set 2:
> it makes sense to cache the failure of the syscall and not attempt to use the
> syscall at every random number generation to then fall back to gnutls
Could you elaborate why that's more efficient?
AFAIK with the way we use getrandom, it can only fail permanently iff we use it
i
Patch Set 2:
I would argue to activate the --enable-gnutls by default, if it is present on
the system. But I don't have a strong opinion here. Or maybe rather, fail
configure.ac if gnutls is not found AND no --disable-gnutls was present.
Furthermore, in terms of efficiency (we allocate TMSIs
Hello Jenkins Builder,
I'd like you to reexamine a change. Please visit
https://gerrit.osmocom.org/4593
to look at the new patch set (#2).
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 8 "jessie" for
example) the osmo_get_rand_id() would always return fai
Review at https://gerrit.osmocom.org/4593
Enable GnuTLS fallback
On systems with GNU/Linux kernel older than 3.17 (Debian 7 "wheezy" for
example) the osmo_get_rand_id() would always return failure. To support
such systems, let's add fallback code which uses GnuTLS library. It have
to be enabled
21 matches
Mail list logo