Re: old top and new -current: missing arcstat sysctl
On Tue, 28 Aug 2018 at 22:46, Alexander Leidinger wrote: > > > > IIRC that would have been the inverted case of running a newer top(1) > > with an older kernel lacking the v_laundry_pages sysctl. In general I'd > > expect us to support running an older top(1) with newer kernels if we > > don't have to bend over backwards to provide compatibility. > > If the new top is summing the 3 up anyway, it sounds like we could > provide the old one as backwards compatibility, even if it is > redundant. I rather have an redundant counter and an old top working > (in the generic case of what we promise to our users; in this specific > case for me I just need to get around to update the jails on the > corresponding systems), than bailing out without displaying anything. I'd support this but don't feel like it justifies breaking the freeze. After the 12.x freeze is over I'll implement this. -- Eitan Adler ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: old top and new -current: missing arcstat sysctl
Quoting Mark Johnston (from Tue, 28 Aug 2018 10:48:42 -0400): On Tue, Aug 28, 2018 at 10:25:39AM -0400, Allan Jude wrote: On 2018-08-28 02:40, Alexander Leidinger wrote: > Hi, > > top reports missing sysctl kstat.zfs.misc.arcstats.other_size for > 12.0-alpha3 with a top from an old-ish -current. > > Is/will this be handled via a compat-11 sysctl (my kernel is without > compat-xx), or did this slip through? That is not something that a compat-xx package can handle. s/package/kernel option/? Sorry, the COMPAT_FREEBSDx kernel options was what I had in mind when I wrote this. That arcstat was broken up into 3 individual stats, which the 12.0-alpha3 version of top will sum together for you. I don't think we've had compat shims like this for previous versions of top, I recall having similar issues when the 'laundry' counter was introduced. IIRC that would have been the inverted case of running a newer top(1) with an older kernel lacking the v_laundry_pages sysctl. In general I'd expect us to support running an older top(1) with newer kernels if we don't have to bend over backwards to provide compatibility. If the new top is summing the 3 up anyway, it sounds like we could provide the old one as backwards compatibility, even if it is redundant. I rather have an redundant counter and an old top working (in the generic case of what we promise to our users; in this specific case for me I just need to get around to update the jails on the corresponding systems), than bailing out without displaying anything. Bye, Alexander. -- http://www.Leidinger.net alexan...@leidinger.net: PGP 0x8F31830F9F2772BF http://www.FreeBSD.orgnetch...@freebsd.org : PGP 0x8F31830F9F2772BF ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: ALPHA3 niggles
On 28/8/18 11:45 pm, Emmanuel Vadot wrote: > On Tue, 28 Aug 2018 22:10:18 +1000 > Brian Scott wrote: > >> Hi, >> >> Just a couple of small observations with 12.0-ALPHA3 from testing on a >> Raspberry-Pi 3: >> >> * The boot loader (I presume the new LUA based one) is sending escape >> sequences to the screen to format up a boot menu. The screen doesn't >> recognise them and shows a series of ^[ style sequences instead. > If it's stuff like that : > https://people.freebsd.org/~manu/RPI2-HDMI.jpg it's known, we should > remove the beastie menu from the image I think Very similar: https://www.dropbox.com/s/0w4buscnk2ec7ez/2018-08-29%2011.56.51.jpg?dl=0 This is using an HDMI attached monitor. I disconnected the serial console adapter last week and so haven't seen what it looks like over that. I'm a bit of a fan of the beastie menu as a general rule so it would be sad to lose it. Still, if the EFI console can't handle the instructions I'm thinking that some simplified version of menu would have to suffice. > >> * The timeout countdown of the boot loader has reverted to very slow. >> The loader did this earlier in the year on -CURRENT but had obviously >> been fixed for a few months now. Maybe some change needs to be adapted >> from the old to the new loader. > I don't have this problem on my RPI3B+ with latest ALPHA snapshot. > The patch is still in the u-boot-rpi3 port > (https://svnweb.freebsd.org/ports/head/sysutils/u-boot-rpi3/files/patch-lib_efi__loader_efi__console.c?revision=468630=markup) > so I don't know what going on for you. One second on the timer takes about 5 seconds of wall clock time. As I said, it had been good for a few months now but was slow earlier in the year. This is the freebsd loader not u-boot unless I completely misunderstand the booting process (also possible), although I suppose u-boot is still providing the console services. The u-boot countdown timer appears to move quickly enough that I haven't really paid any attention to it. > >> * The root user on the image doesn't have a .login file. > The users are added with pw directly (iirc) and the /etc/skel content > isn't copied in their home directory, that's something we should change > for 13 and mfc back for 12.1 > >> I realise these are all cosmetic issues in the overall scheme of things >> but if anyone is working on a bit of spit-and-polish for the new release >> they might want to have a look at a few of these. >> >> Otherwise I'm happy to say that I haven't found any significant problems >> yet. I'll keep looking though. >> >> Keep up the good work, >> >> Brian >> >> ___ >> freebsd-current@freebsd.org mailing list >> https://lists.freebsd.org/mailman/listinfo/freebsd-current >> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" Cheers, Brian ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: missing AMQ usage in dummynet/ipfw man page for 12-alpha3
Got it, thanks! Rodney W. Grimes 于2018年8月28日周二 下午9:40写道: > > hi, I see the AMQ function is there but there is no usage info. > > And find this review marked accepted: > > Dummynet AQM usage documentation for ipfw man page ? > > https://reviews.freebsd.org/D12507 > > That review is accepted by manpages, it has not been > reviewed outside of that, and is in my inbasket, but > not of extremly high priority. > > It has not been commited to the tree. > > -- > Rod Grimes > rgri...@freebsd.org > ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Current and nvidia-driver again
On 28/08/2018 10:14, Alex V. Petrov wrote: /usr/src Revision: 338342 ===> Building for nvidia-driver-390.77 Hi, Have you tried applying the patch at https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=230780 ? Update your ports tree first. Apply it in /usr/ports It worked for me on r338323 -- J. ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: old top and new -current: missing arcstat sysctl
On Tue, Aug 28, 2018 at 10:25:39AM -0400, Allan Jude wrote: > On 2018-08-28 02:40, Alexander Leidinger wrote: > > Hi, > > > > top reports missing sysctl kstat.zfs.misc.arcstats.other_size for > > 12.0-alpha3 with a top from an old-ish -current. > > > > Is/will this be handled via a compat-11 sysctl (my kernel is without > > compat-xx), or did this slip through? > > > > Bye, > > Alexander. > > > > That is not something that a compat-xx package can handle. s/package/kernel option/? > That arcstat was broken up into 3 individual stats, which the > 12.0-alpha3 version of top will sum together for you. > > I don't think we've had compat shims like this for previous versions of > top, I recall having similar issues when the 'laundry' counter was > introduced. IIRC that would have been the inverted case of running a newer top(1) with an older kernel lacking the v_laundry_pages sysctl. In general I'd expect us to support running an older top(1) with newer kernels if we don't have to bend over backwards to provide compatibility. ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: old top and new -current: missing arcstat sysctl
On 2018-08-28 02:40, Alexander Leidinger wrote: > Hi, > > top reports missing sysctl kstat.zfs.misc.arcstats.other_size for > 12.0-alpha3 with a top from an old-ish -current. > > Is/will this be handled via a compat-11 sysctl (my kernel is without > compat-xx), or did this slip through? > > Bye, > Alexander. > That is not something that a compat-xx package can handle. That arcstat was broken up into 3 individual stats, which the 12.0-alpha3 version of top will sum together for you. I don't think we've had compat shims like this for previous versions of top, I recall having similar issues when the 'laundry' counter was introduced. -- Allan Jude signature.asc Description: OpenPGP digital signature
Re: ALPHA3 niggles
> Hi, > > Just a couple of small observations with 12.0-ALPHA3 from testing on a > Raspberry-Pi 3: > > *??? The boot loader (I presume the new LUA based one) is sending escape > sequences to the screen to format up a boot menu. The screen doesn't > recognise them and shows a series of ^[ style sequences instead. > > *??? The timeout countdown of the boot loader has reverted to very slow. > The loader did this earlier in the year on -CURRENT but had obviously > been fixed for a few months now. Maybe some change needs to be adapted > from the old to the new loader. > > *??? The root user on the image doesn't have a .login file. Confirmed on an amd64 install, adding brd@ to cc list, this is an issue for him. > I realise these are all cosmetic issues in the overall scheme of things > but if anyone is working on a bit of spit-and-polish for the new release > they might want to have a look at a few of these. > > Otherwise I'm happy to say that I haven't found any significant problems > yet. I'll keep looking though. > > Keep up the good work, > > Brian > > ___ > freebsd-current@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" > > -- Rod Grimes rgri...@freebsd.org ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: ALPHA3 niggles
On Tue, 28 Aug 2018 22:10:18 +1000 Brian Scott wrote: > Hi, > > Just a couple of small observations with 12.0-ALPHA3 from testing on a > Raspberry-Pi 3: > > * The boot loader (I presume the new LUA based one) is sending escape > sequences to the screen to format up a boot menu. The screen doesn't > recognise them and shows a series of ^[ style sequences instead. If it's stuff like that : https://people.freebsd.org/~manu/RPI2-HDMI.jpg it's known, we should remove the beastie menu from the image I think > * The timeout countdown of the boot loader has reverted to very slow. > The loader did this earlier in the year on -CURRENT but had obviously > been fixed for a few months now. Maybe some change needs to be adapted > from the old to the new loader. I don't have this problem on my RPI3B+ with latest ALPHA snapshot. The patch is still in the u-boot-rpi3 port (https://svnweb.freebsd.org/ports/head/sysutils/u-boot-rpi3/files/patch-lib_efi__loader_efi__console.c?revision=468630=markup) so I don't know what going on for you. > * The root user on the image doesn't have a .login file. The users are added with pw directly (iirc) and the /etc/skel content isn't copied in their home directory, that's something we should change for 13 and mfc back for 12.1 > I realise these are all cosmetic issues in the overall scheme of things > but if anyone is working on a bit of spit-and-polish for the new release > they might want to have a look at a few of these. > > Otherwise I'm happy to say that I haven't found any significant problems > yet. I'll keep looking though. > > Keep up the good work, > > Brian > > ___ > freebsd-current@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" -- Emmanuel Vadot ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: missing AMQ usage in dummynet/ipfw man page for 12-alpha3
> hi, I see the AMQ function is there but there is no usage info. > And find this review marked accepted: > Dummynet AQM usage documentation for ipfw man page ? > https://reviews.freebsd.org/D12507 That review is accepted by manpages, it has not been reviewed outside of that, and is in my inbasket, but not of extremly high priority. It has not been commited to the tree. -- Rod Grimes rgri...@freebsd.org ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: ALPHA3 niggles
On Tue, Aug 28, 2018 at 7:10 AM Brian Scott wrote: > > Hi, > > Just a couple of small observations with 12.0-ALPHA3 from testing on a > Raspberry-Pi 3: > > *The boot loader (I presume the new LUA based one) is sending escape > sequences to the screen to format up a boot menu. The screen doesn't > recognise them and shows a series of ^[ style sequences instead. What terminal are you using here? > *The timeout countdown of the boot loader has reverted to very slow. > The loader did this earlier in the year on -CURRENT but had obviously > been fixed for a few months now. Maybe some change needs to be adapted > from the old to the new loader. When you say "very slow" -- your autoboot counter is defaulting high, or there's a perceptible longer-than-one-second delay between updates? Thanks, Kyle Evans ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
ALPHA3 niggles
Hi, Just a couple of small observations with 12.0-ALPHA3 from testing on a Raspberry-Pi 3: * The boot loader (I presume the new LUA based one) is sending escape sequences to the screen to format up a boot menu. The screen doesn't recognise them and shows a series of ^[ style sequences instead. * The timeout countdown of the boot loader has reverted to very slow. The loader did this earlier in the year on -CURRENT but had obviously been fixed for a few months now. Maybe some change needs to be adapted from the old to the new loader. * The root user on the image doesn't have a .login file. I realise these are all cosmetic issues in the overall scheme of things but if anyone is working on a bit of spit-and-polish for the new release they might want to have a look at a few of these. Otherwise I'm happy to say that I haven't found any significant problems yet. I'll keep looking though. Keep up the good work, Brian ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Current and nvidia-driver again
/usr/src Revision: 338342 ===> Building for nvidia-driver-390.77 nvidia_subr.c:1179:65: error: too many arguments to function call, expected 2, have 3 kmem_free(kmem_arena, at->pte_array[0].virtual_address, at->size); ~ ^~~~ /usr/src/sys/vm/vm_extern.h:69:1: note: 'kmem_free' declared here void kmem_free(vm_offset_t addr, vm_size_t size); ^ nvidia_subr.c:1205:61: error: too many arguments to function call, expected 2, have 3 kmem_free(kmem_arena, at->pte_array[0].virtual_address, at->size); ~ ^~~~ /usr/src/sys/vm/vm_extern.h:69:1: note: 'kmem_free' declared here void kmem_free(vm_offset_t addr, vm_size_t size); ^ nvidia_subr.c:1322:65: error: too many arguments to function call, expected 2, have 3 kmem_free(kmem_arena, at->pte_array[i].virtual_address, PAGE_SIZE); ~ ^ ./machine/param.h:101:19: note: expanded from macro 'PAGE_SIZE' #define PAGE_SIZE (1
Re: drm / drm2 removal in 12
blubee blubeeme writes: > You seem to miss the point where the you avoid breaking the system for > any users not on the bleeding edge. You seem to miss the point where nobody is interested in anything you have to say any more. DES -- Dag-Erling Smørgrav - d...@des.no ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
missing AMQ usage in dummynet/ipfw man page for 12-alpha3
hi, I see the AMQ function is there but there is no usage info. And find this review marked accepted: Dummynet AQM usage documentation for ipfw man page : https://reviews.freebsd.org/D12507 ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
old top and new -current: missing arcstat sysctl
Hi, top reports missing sysctl kstat.zfs.misc.arcstats.other_size for 12.0-alpha3 with a top from an old-ish -current. Is/will this be handled via a compat-11 sysctl (my kernel is without compat-xx), or did this slip through? Bye, Alexander. -- http://www.Leidinger.net alexan...@leidinger.net: PGP 0x8F31830F9F2772BF http://www.FreeBSD.orgnetch...@freebsd.org : PGP 0x8F31830F9F2772BF ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"