Date: Fri, 28 May 2021 21:49:50 -0700 From: Mark Millard via freebsd-ppc <freebsd-ppc@freebsd.org> To: Jason Bacon <bacon4000@gmail.com> Cc: "freebsd-ppc@freebsd.org" <freebsd-ppc@freebsd.org> Subject: Re: 13.0 packages? Message-ID: <FE25B1E8-9D46-4162-93C8-D3F219DA7F8E@yahoo.com> In-Reply-To: <200faa1c-6153-1e2f-d3a1-f76eaf83284a@gmail.com> References: <b4cbbffd-1d95-ff4e-2608-9633581dadd4@gmail.com> <9FF7E0E7-E666-443C-8CA7-C031EBE60BF6@yahoo.com> <200faa1c-6153-1e2f-d3a1-f76eaf83284a@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2021-May-28, at 17:52, Jason Bacon <bacon4000 at gmail.com> wrote: > On 5/26/21 9:40 PM, Mark Millard wrote: >> On 2021-May-20, at 17:26, Jason Bacon <bacon4000 at gmail.com> wrote: >>> What's the plan for 13.0 powerpc64 packages? >>>=20 >>> I have a Mac G5 I was using for port testing, but since upgrading to = 13.0-RELEASE, there haven't been any binary packages for it. Building = everything from source takes too long. >> https://pkg.freebsd.org has a link to: >> =E2=80=A2 FreeBSD:13:powerpc64 ( = https://pkg.freebsd.org/FreeBSD%3A13%3Apowerpc64 ) >> There the dates on things indicate that quarterly is newer >> than latest. meta.txz is reported as having date/time: >> 2021-May-20 07:07 >> (as seen in the time zone I'm in). >> https://pkg-status.freebsd.org shows that jail 130powerpc64 started >> building on Tue, 18 May 2021 21:01:05 GMT and reached stopped:done: >> status, taking an elapsed time of 34:06:48 . It also reports that >> the server was: pylon . >> But the linked page: >> = https://pkg-status.freebsd.org/builds/default:quarterly:130powerpc64:cdf5b= fe36fe5:pylon >> reports only 1976 ports built of the only 4340 that were queued >> for that specific build. >> There was another jail that built far more: 130releng-powerpc64 is >> indicated to have built 23522 of the 30700 queued. It is listed >> as started on Sun, 04 Apr 2021 13:48:51 GMT, having reached a >> status of stopped:crashed:, with an elapsed time of 73:28:43 . >> Also pylon. >> It appears that http://pylon.nyi.freebsd.org is not currently >> accessible. This limits what can be investigated. >> I would guess that: >> = https://pkg.freebsd.org/FreeBSD:13:powerpc64/quarterly/packagesite.txz >> would give a clue what should be accessible via quarterly. >> It is listed as being 6058988 Bytes in size. >> =3D=3D=3D >> Mark Millard >> marklmi at yahoo.com >> ( dsl-only.net went >> away in early 2018-Mar) >=20 > Thanks for the status update. I'm actually more curious about the = intentions for the future, though. Are there plans to build current = and/or quarterly packages for 13.0 on a regular rotation? Is it just a = matter of manpower and hardware to get it rolling? I've no direct knowledge. My only clue is evidence left behind. The existing pylon package-builds include recent attempts (this month) and older ones: git context: quarterly 130powerpc64 Tue, 18 May 2021 21:01:05 GMT 34:06:48 quarterly 122powerpc64 Tue, 18 May 2021 02:34:32 GMT 18:26:26 default main-powerpc64 Sat, 08 May 2021 08:16:21 GMT 42:38:52 svn context: quarterly 130releng-powerpc64 Sun, 04 Apr 2021 13:48:51 GMT 73:57:51 quarterly 121powerpc64 Sat, 23 Jan 2021 02:59:07 GMT 73:28:43 default head-powerpc64 Tue, 22 Dec 2020 18:07:32 GMT 157:13:50 default 120powerpc64 Tue, 24 Mar 2020 21:37:02 GMT 01:25:55 That is all that https://pkg-status.freebsd.org/builds?server=3Dpylon shows. It suggests specific single attempts to build: quarterly for 13.0 and 12.2 default [so: latest] for main [so: 14] as well as a specific quarterly to go with release/13.0.0 's distribution (built in early April). (I ignore older here.) But it also makes clear that there are fairly long periods between attempts, so regular builds of updated ports have not been part of the activity so far. So far as I know, all the above are Big Endian for the context and target, no little endian jails (yet?). pylon does seem to still be inaccessible (for looking at build logs and the like). > Will support for big-endian wane as the focus shifts to LE? I'm not aware of evidence left behind beyond the absence of any little endian jails in the list. If it goes that direction, it likely would take some time for little endian to become primary. FYI: With old PowerMac failures and other issues, my powerpc64 and 32-bit powerpc activity has largely stopped: it was only PowerMac based. =3D=3D=3D Mark Millard marklmi at yahoo.com ( dsl-only.net went away in early 2018-Mar)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?FE25B1E8-9D46-4162-93C8-D3F219DA7F8E>