On 5 October 2015 at 11:51, Josh Elsasser <j...@elsasser.org> wrote: > On Mon, Oct 05, 2015 at 08:59:45AM -0400, Kenneth Westerback wrote: >> On 2 October 2015 at 08:27, Kenneth Westerback <kwesterb...@gmail.com> wrote: >> > On 2 October 2015 at 08:18, Jérémie Courrèges-Anglas <j...@wxcvbn.org> >> > wrote: >> >> Kenneth Westerback <kwesterb...@gmail.com> writes: >> >> >> >>> On 30 September 2015 at 04:11, Jérémie Courrèges-Anglas >> >>> <j...@wxcvbn.org> wrote: >> >>>> >> >>>> Last night I ran ''make; make clean'' in a loop on i386 and amd64. >> >>>> >> >>>> amd64: 142 builds / 0 failures >> >>>> i386: 58 builds / 2 failures >> >>>> >> >>>> (same failure as yours) >> >>>> >> >>>> Maybe this has something to do with this... >> >>>> >> >>>> [...] >> >>>> checking for the code address range... 0x16000000 >> >>>> checking for the malloc address range... 0x87000000 >> >>>> checking for the shared library address range... 0x2D000000 >> >>>> checking for the stack address range... 0xCF000000 >> >>>> [...] >> >>>> >> >>>> -- >> >>>> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 >> >>>> E7EE >> >>> >> >>> Black magic as far as I am concerned. :-) >> >> >> >> 2.48 fails at the same rate on i386. >> >> >> >> -- >> >> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 >> >> E7EE >> > >> > In which case I would consider it fine to upgrade. :-) >> > >> > My sparc64 is resisting me but I hope to overcome it's reluctance >> > today and start a build on -current. >> > >> > .... Ken >> >> It's still resisting. I am getting worried about it. :-( >> >> .... Ken > > I thought I'd jumped in here but it looks like not. > > This builds and appears to pass tests on macppc, and builds an sbcl > which also passes its tests. > > The fact that the configure script appears to be discovering and > hardcoding random memory addresses is concerning, but I guess it was > doing that already. > > Anyway, I'm OK with this and please feel free to remove me as > maintainer and add yourself. My only interest in clisp is for > bootstrapping sbcl anyway.
Now I just feel bad. :-(. I'm ok with this going in as well unless you want to wait for a sparc64 test. .... Ken
