Back to Home / #xen / 2022 / 04 / Prev Day | Next Day
#xen IRC Logs for 2022-04-14

---Logopened Thu Apr 14 00:00:36 2022
00:19-!-ChmEarl [~prymar56@0002b86c.user.oftc.net] has quit [Quit: Leaving]
02:30-!-jgross_ is now known as jgross
04:54-!-overholts [~overholts@129.159.36.225] has quit [Ping timeout: 480 seconds]
09:07<weylkesiq>there is a 4.16.1 release available, but i'm not seeing it announced anywhere...is this intentional?
09:07<weylkesiq>i'm hitting an error while building it, specifically:
09:08<weylkesiq>bunzip2.c: In function ‘get_next_block’:
09:08<weylkesiq>bunzip2.c:261:41: error: ‘length’ may be used uninitialized [-Werror=maybe-uninitialized]
09:08<weylkesiq>261 minLen = maxLen = length[0];
09:09<gwd>weylkesiq: It will be announced soon.
09:09<gwd>There's a lot of machinery, much of which is manual and done my different people
09:10<weylkesiq>gwd: i see, do you have any idea why i'm hitting that error? nothing relevant seems to have happened to bunzip2.c recently
09:10<gwd>weylkesiq: No; ideally ask on xen-devel@lists.xenproject.org, so anyone doing a google search will find your question and (hopefully) the answer
09:10<royger>weylkesiq: which compiler are you using?
09:11<weylkesiq>royger: gcc (Gentoo Hardened 11.2.1_p20220115 p4) 11.2.1 20220115
09:12<weylkesiq>and that hasn't changed since i last built xen
09:12<gwd>Normally that sort of thing happens when a new version of the compiler notices something dodgy that it didn't notice before
09:13<weylkesiq>binutils and glibc have been bumped though
09:15<royger>so you didn't hit that when building 4.15 but hit it now with 4.16?
09:15<royger>sorry, 4.16.0 vs 4.16.1
09:15<andyhhp>binutils and glibc won't make a difference here. It's down to the compiler, and any code changes
09:16<weylkesiq>royger: correct, 4.16 worked
09:16<andyhhp>there were changes in bunzip2.c for CET-IBT, which GCC 11.2 does support
09:17<weylkesiq>4.16.0 i should say
09:17<andyhhp>but I don't see how they'd plausibly change this
09:17<royger>did we change some cflags?
09:18<andyhhp>well by default, we'll start using -mcf-harden=branch
09:20<gwd>I mean, the most important question si whether the warning is correct..
09:21<andyhhp>actually the warning does come from GCC's analysis of the AST, but it is a timebound search because it easily turns into exponential complexity
09:21<andyhhp>lemme stare at some source
09:22<andyhhp>but also, why wasn't this picked up in Gitlab CI? (Rhetorical question. The two answers are a) we've not got any GCC 11 in there, and b) it's still post-push and only when people go looking)
09:23<royger>we do symCount = foo + 2, and then for (i = 0; i < symCount; i++) { length[i] = ... }
09:24<royger>i is not unsigned however, so under weird circumstances we could end up not iterating? (I would expect things will likely explode elsewhere if this happens)
09:24<royger>Linux has the same logic AFAICT
09:24<andyhhp>right - there were no actual code changes in bunzip2.c between 4.16.0 and 4.16.1
09:25<andyhhp>so it will be some side effect of a change in the compiler, or cflags
09:34<weylkesiq>hm, i may be missing some patches that were applied to 4.16.0
09:42<weylkesiq>aha!
09:42<weylkesiq>-CFLAGS += -Werror -Wredundant-decls -Wno-pointer-arith
09:42<weylkesiq>+CFLAGS += -Wredundant-decls -Wno-pointer-arith
09:42<weylkesiq>the ebuild is confusing and i was missing that
09:42<weylkesiq>that's some pretty heavy-handed patching though...
09:59<weylkesiq>so applied the missing patches and build went through, sorry for the noise
10:06<royger>well, we need to fix this upstream ragardless
10:13<andyhhp>eww - I've just looked at Gentoo's ebuild script for Xen. That is in desperate need of some care an attention
10:35<weylkesiq>andyhhp: quite right, it's not in good shape
10:36<weylkesiq>but 4.16.1 booted and tested! and i'm happy to see that someone has fixed my problems with dom0 PVH in the 5.17 kernel
10:52<ccw>Is there any way that a domU guest can query it's domain name?
10:53<andyhhp>`xenstore-read name`
10:55<ccw>ah! .. so the xen tools can be installed in the guest..
10:56<andyhhp>xenstore in particular can and commonly are
10:56<ccw>I've been using xen for quite a while and didn't realize that..
11:13-!-ChmEarl [~prymar56@098-147-150-167.res.spectrum.com] has joined #xen
11:13-!-ChmEarl is "Mark Pryor" on #xen ##xen-packaging #mock #packaging #virt
11:52-!-ccw [~ccw@0002a6fe.user.oftc.net] has quit [Quit: ZNC - https://znc.in]
11:52-!-ccw [~ccw@97-122-67-169.hlrn.qwest.net] has joined #xen
11:52-!-ccw is "Cheyenne Wills" on #xen #virt #bpftrace
21:46-!-ChmEarl [~prymar56@0002b86c.user.oftc.net] has quit [Quit: Leaving]
22:23-!-jos2 [~jos3@dyndsl-091-096-043-100.ewe-ip-backbone.de] has joined #xen
22:23-!-jos2 is "jos4" on #xen #virt #Qubes_OS #kvm #oftc
22:29-!-jgross_ [~juergen_g@2a01:41e1:2b7a:c800:7c2e:a90d:65b1:cfd0] has joined #xen
22:29-!-jgross_ is "realname" on #xen
22:31-!-jos1 [~jos3@host-091-097-187-151.ewe-ip-backbone.de] has quit [Ping timeout: 480 seconds]
22:36-!-jgross [~juergen_g@2a01:41e1:2b4a:6f00:6e2c:f85a:7e40:591e] has quit [Ping timeout: 480 seconds]
---Logclosed Fri Apr 15 00:00:37 2022