Re: [gentoo-user] updating remote system
On Tue, 2023-04-11 at 20:24 -0400, jul...@jroy.ca wrote: > On Tue, 2023-04-11 at 18:21 -0600, the...@sys-concept.com wrote: > > What is the correct way of updating remote system? > > I'll need to update the kernel + world > > > > 1.) Update all packages in world first, reboot and update kernel > > 2.) Update kernel first, reboot and update world > > > > It seems to me logical choice is 2.) as I'll need new kernel to > > update to new VirtualBox. > > Right now remote system is running linux-5.10.103-gentoo and I > > think > > new VirtualBox need kerenl-5.13 or higher. > > > > > > -- It seems my message got sent empty... Let's try again: I usually update world first, then the kernel, and finally reboot. -- Julien signature.asc Description: This is a digitally signed message part
Re: [gentoo-user] updating remote system
On Tue, 2023-04-11 at 18:21 -0600, the...@sys-concept.com wrote: > What is the correct way of updating remote system? > I'll need to update the kernel + world > > 1.) Update all packages in world first, reboot and update kernel > 2.) Update kernel first, reboot and update world > > It seems to me logical choice is 2.) as I'll need new kernel to > update to new VirtualBox. > Right now remote system is running linux-5.10.103-gentoo and I think > new VirtualBox need kerenl-5.13 or higher. > -- Julien signature.asc Description: This is a digitally signed message part
[gentoo-user] updating remote system
What is the correct way of updating remote system? I'll need to update the kernel + world 1.) Update all packages in world first, reboot and update kernel 2.) Update kernel first, reboot and update world It seems to me logical choice is 2.) as I'll need new kernel to update to new VirtualBox. Right now remote system is running linux-5.10.103-gentoo and I think new VirtualBox need kerenl-5.13 or higher. -- Thelma
Re: [gentoo-user] X not starting after kernel upgrade
On 4/11/23 09:47, Jack wrote: [snip] message after recompiling Nvida driver: WARN: postinst Be warned/reminded that the 390.xx branch reached end-of-life and NVIDIA is no longer fixing issues (including security). Free to keep using (for now) but it is recommended to either switch to nouveau or replace hardware. Will be kept in-tree while possible, but expect it to be removed likely in early 2027 or earlier if major issues arise. It looks like the 390 driver has been masked in portage, although not yet removed. Have you read the message in /usr/portage/profiles/package.mask? As the message suggests, you may be better off switching to actually using nouveau. It also implies that 6.1.X LTS kernels "should" still work. Note that the X log you posted doesn't even show it trying to load the nvidia driver - have you tried again after recompiling it for the new kernel? That message was a warning, not an error, so it should have compiled successfully - did it? Good suggestion, I switched to nouveau and was as able to upgrade to newest kernel-6.1.19 as well. besides, nvidia-390 is on the end of line as it might be removed from portage one day.
Re: [gentoo-user] xfce4 automount usb doesn't work
On 4/11/23 14:37, Matt Connell wrote: On Tue, 2023-04-11 at 11:28 -0600, the...@sys-concept.com wrote: Does user need to be in group: plugdev for it to work. I'm fairly confident that you do, yes. "Udisks uses polkit to handle permissions. Make sure each user is in the plugdev group" https://wiki.gentoo.org/wiki/Udisks#Configuration None of my systems have even acct-group/plugdev installed and no user is in plugdev group Two of the system mounts USB automatically (correctly) when USB is plugged-in, one updated last Dec. and one yesterday. Third computer upgraded yesterday but USB doesn't auto-mount automatically. One of my remote computer stop automounting USB all of a sudden as well, even though I did not change any setting it happen after I installed gnome-extra/zenity
[gentoo-user] Re: Portage 'Completed (m of n)' messages
On 11/04/2023 18:26, Peter Humphrey wrote: On Tuesday, 11 April 2023 14:36:27 BST Arve Barsnes wrote: On Tue, 11 Apr 2023 at 14:45, Peter Humphrey wrote: On Tuesday, 11 April 2023 13:28:58 BST Nikos Chantziaras wrote: On 11/04/2023 13:59, Peter Humphrey wrote: What does the panel think of these new status messages from portage (~amd64)? What messages? Where? When? :P They seem to be a feature of sys-apps/portage-3.0.46 and probably 3.0.45. Nikos' questions still stand :P We all know that, when a package is emerged, portage announces the start of the process, and the start of installation to the live disk. Well, now, it also announces completion of all work on the package. That causes another length of introductory string, so now there are three columns to scan for package numbers. It's a personal preference, I'm sure - even GOM(s) don't all agree. :) Oh, I thought there were actual desktop notifications or something, which sounded weird. I almost never look the terminal when emerging :P
Re: [gentoo-user] xfce4 automount usb doesn't work
On Tue, 2023-04-11 at 11:28 -0600, the...@sys-concept.com wrote: > Does user need to be in group: plugdev for it to work. I'm fairly confident that you do, yes. "Udisks uses polkit to handle permissions. Make sure each user is in the plugdev group" https://wiki.gentoo.org/wiki/Udisks#Configuration
Re: [gentoo-user] xfce4 automount usb doesn't work
On 4/11/23 10:15, Matt Connell wrote: On Tue, 2023-04-11 at 09:43 -0600, the...@sys-concept.com wrote: When USB is inserted, the icon appears on a desktop but it is not aouto mounted. In settings: Removable Drive and Media --> - Mount removable drive when hot-plugged (is checked) Does it have something to do with dbus? My first guess is permissions. Can you mount the drive in Thunar by clicking on it in the side panel? Does user need to be in group: plugdev for it to work. On my other older system USB automount works correctly and I'm not in a group plugdev
Re: [gentoo-user] xfce4 automount usb doesn't work
On 4/11/23 10:15, Matt Connell wrote: On Tue, 2023-04-11 at 09:43 -0600, the...@sys-concept.com wrote: When USB is inserted, the icon appears on a desktop but it is not aouto mounted. In settings: Removable Drive and Media --> - Mount removable drive when hot-plugged (is checked) Does it have something to do with dbus? My first guess is permissions. Can you mount the drive in Thunar by clicking on it in the side panel? Yes, when the usb-icon appears on desktop, right clicking on it allows me to mount it. On the system that works correctly, when I insert the USB I get: [822766.172299] sda: sda1 [822766.173269] sd 4:0:0:0: [sda] Attached SCSI removable disk [822776.321898] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: errors=remount-ro On the newly updated system I get only: [12095.359150] sdb: sdb1 [12095.359219] sd 9:0:0:0: [sdb] Attached SCSI removable disk
Re: [gentoo-user] xfce4 automount usb doesn't work
On Tue, 2023-04-11 at 09:43 -0600, the...@sys-concept.com wrote: > When USB is inserted, the icon appears on a desktop but it is not aouto > mounted. > > In settings: Removable Drive and Media --> > - Mount removable drive when hot-plugged (is checked) > > Does it have something to do with dbus? > My first guess is permissions. Can you mount the drive in Thunar by clicking on it in the side panel?
Re[2]: [gentoo-user] Portage 'Completed (m of n)' messages
Neil Bothwick wrote: > On Tue, 11 Apr 2023 11:59:08 +0100, Peter Humphrey wrote: >> What does the panel think of these new status messages from portage >> (~amd64)? At first I thought "that's useful", but after a while I >> concluded that it just adds to the clutter on the screen and actually >> impedes my ability to follow progress. > I like it. Previously, I had to wait for the completed counter to > increment before knowing that installation of a package was complete. > It's handy if you need to interrupt an emerge session and don't want to > do it just before Chromium has finished installing :-O >> Perhaps I'm just being a Grumpy Old Man (tm). > See above, some Grumpy Old Men like it ;-) Now we get to see "Emerging package" right before building starts "Installing package" between building and installing and "Completed" after a package is really done. From the viewpoint of this grumpy old guy watching a bunch of servers doing their weekly update dance, the "Completed" line is much more interesting than the "Installing", which most of the time only indicates that the load is switching from CPU to disk I/O. So when will the "Installing" lines go away for good? s. -- Ever wondered how fence fights between pensioners begin?
Re: [gentoo-user] X not starting after kernel upgrade
On 2023.04.10 20:14, the...@sys-concept.com wrote: On 4/10/23 17:29, Jack wrote: On 2023.04.10 18:22, the...@sys-concept.com wrote: On 4/10/23 15:53, the...@sys-concept.com wrote: After upgrading to newest kernel the X will not start. Is: make oldconfig same as: make olddefconfig ? Xorg.0.log showing; [ 673.829] (II) Loading /usr/lib64/xorg/modules/drivers/nouveau_drv.so [ 673.829] (II) Module nouveau: vendor="X.Org Foundation" [ 673.829] compiled for 1.21.1.8, module version = 1.0.17 [ 673.829] Module class: X.Org Video Driver [ 673.829] ABI class: X.Org Video Driver, version 25.2 [ 673.829] (II) LoadModule: "nv" [ 673.829] (WW) Warning, couldn't open module nv [ 673.829] (EE) Failed to load module "nv" (module does not exist, 0) [ 673.829] (II) LoadModule: "modesetting" [ 673.829] (II) Loading /usr/lib64/xorg/modules/drivers/modesetting_drv.so [ 673.830] (II) Module modesetting: vendor="X.Org Foundation" [ 673.830] compiled for 1.21.1.8, module version = 1.21.1 [ 673.830] Module class: X.Org Video Driver [ 673.830] ABI class: X.Org Video Driver, version 25.2 [ 673.830] (II) LoadModule: "fbdev" [ 673.830] (WW) Warning, couldn't open module fbdev [ 673.830] (EE) Failed to load module "fbdev" (module does not exist, 0) [ 673.830] (II) LoadModule: "vesa" [ 673.830] (WW) Warning, couldn't open module vesa [ 673.830] (EE) Failed to load module "vesa" (module does not exist, 0) [ 673.830] (II) NOUVEAU driver [ 673.830] (II) NOUVEAU driver for NVIDIA chipset families : [ 673.830] RIVA TNT (NV04) [ 673.830] RIVA TNT2 (NV05) [ 673.830] GeForce 256 (NV10) [ 673.830] GeForce 2 (NV11, NV15) [ 673.830] GeForce 4MX (NV17, NV18) [ 673.830] GeForce 3 (NV20) [ 673.830] GeForce 4Ti (NV25, NV28) [ 673.830] GeForce FX (NV3x) [ 673.830] GeForce 6 (NV4x) [ 673.830] GeForce 7 (G7x) [ 673.830] GeForce 8 (G8x) [ 673.830] GeForce 9 (G9x) [ 673.830] GeForce GTX 2xx/3xx (GT2xx) [ 673.830] GeForce GTX 4xx/5xx (GFxxx) [ 673.830] GeForce GTX 6xx/7xx (GKxxx) [ 673.830] GeForce GTX 9xx (GMxxx) [ 673.830] GeForce GTX 10xx (GPxxx) [ 673.830] (II) modesetting: Driver for Modesetting Kernel Drivers: kms [ 673.831] (EE) [drm] Failed to open DRM device for pci::01:00.0: -19 [ 673.831] (EE) open /dev/dri/card0: No such file or directory [ 673.831] (WW) Falling back to old probe method for modesetting [ 673.831] (EE) open /dev/dri/card0: No such file or directory [ 673.831] (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support [ 673.831] (EE) Screen 0 deleted because of no matching config section. [ 673.831] (II) UnloadModule: "modesetting" [ 673.831] (EE) Device(s) detected, but none match those in the config file. [ 673.831] (EE) Fatal server error: [ 673.831] (EE) no screens found(EE) This is an old system and is using: GeForce GTS 450 Does the newest kernel: linux-6.1.19-gentoo supports my old GeForce GTS 450 card? grep KMS .config CONFIG_DRM_KMS_HELPER=y CONFIG_DRM=y I don't recall anything about the kernel that has to match the video card - it's the video driver for X, but if you're using Nouveau, I wouldn't expect a problem. I know the current nVidia driver is often not ready for the latest kernel, but I don't think that's your problem. Have you recompiled xv86-video-nouveau after installing the new kernel? Are there any relevant errors in dmesg? I think the problem is with latest kernel linux-6.1.19-gentoo I'm using Nvidia dirver; x11-drivers/nvidia-drivers-390.157 and after compiling the driver on this kernel I was getting a message notice: WARN: setup Detected potential configuration issues witDIVIDEND RECEIVEDh used kernel: CONFIG_FB_SIMPLE: is set, recommended to disable and switch to FB_EFI or FB_VESA as it currently may be broken with >=kernel-5.18.13 + NVIDIA: https://github.com/NVIDIA/open-gpu-kernel-modules/issues/341 (feel free to ignore this if it works for you) I downgraded the kerenl to: linux-5.15.102-gentoo and X is booting OK. message after recompiling Nvida driver: WARN: postinst Be warned/reminded that the 390.xx branch reached end-of-life and NVIDIA is no longer fixing issues (including security). Free to keep using (for now) but it is recommended to either switch to nouveau or replace hardware. Will be kept in-tree while possible, but expect it to be removed likely in early 2027 or earlier if major issues arise. It looks like the 390 driver has been masked in portage, although not yet removed. Have you read the message in /usr/portage/profiles/package.mask? As the message suggests, you may be better off switching to actually using nouveau. It also implies that 6.1.X LTS kernels "should"
[gentoo-user] xfce4 automount usb doesn't work
When USB is inserted, the icon appears on a desktop but it is not aouto mounted. In settings: Removable Drive and Media --> - Mount removable drive when hot-plugged (is checked) Does it have something to do with dbus? -- Thelma
Re: [gentoo-user] config file '/etc/mtab' needs updating
On 2023.04.11 09:14, Matt Connell wrote: On Mon, 2023-04-10 at 23:44 -0600, the...@sys-concept.com wrote: > After update I get: > * IMPORTANT: config file '/etc/mtab' needs updating. > > What is this, don't remember seeing it before. > > cfg-update -u > doesn't give me an option to view it. > > dispatch-conf will show you what is being changed and give you the option to use/zap the change. I never even knew cfg-update existed (I've always used dispatch-conf). I've been using etc-update. It has an interactive merge feature (only two-way, not three) and I sometimes do the merge manually in emacs, which has a three-way merge, although I have not used it.
Re: [gentoo-user] emerge -U or emerge -N
On Tue, 11 Apr 2023 10:18:13 -0500, Dale wrote: > >> When I say deeper, I mean it will find more packages that may not be > >> found otherwise. > > And -e finds even more - but more is not always better. -U was > > introduced because -N was causing too many packages to be rebuilt > > unnecessarily. > Before those options came along, I would run emerge -e world to fix > problems. Sometimes revdep-rebuild would catch things but sometimes it > wouldn't. Thing is, since I started using the current options, I have > few problems with package upgrades. Sure, they have a known bug on > occasion but recompiling won't help that. I'm just talking about > problems with one package not matching up with some other package and > recompile fixes it. > > To me, the fact it works so much better tells me I'm doing something > right. Other people may get different results but as long as what I'm > doing works, I don't plan to change anything. -N works, it just creates extra work. If you are happy with that, there's no need to change. -- Neil Bothwick There are two hard things in computer science: cache invalidation, naming things and off-by-one errors. pgp3tqCYjPe6V.pgp Description: OpenPGP digital signature
Re: [gentoo-user] Re: Portage 'Completed (m of n)' messages
On Tuesday, 11 April 2023 14:36:27 BST Arve Barsnes wrote: > On Tue, 11 Apr 2023 at 14:45, Peter Humphrey wrote: > > On Tuesday, 11 April 2023 13:28:58 BST Nikos Chantziaras wrote: > > > On 11/04/2023 13:59, Peter Humphrey wrote: > > > > What does the panel think of these new status messages from portage > > > > (~amd64)? > > > > > > What messages? Where? When? :P > > > > They seem to be a feature of sys-apps/portage-3.0.46 and probably 3.0.45. > > Nikos' questions still stand :P We all know that, when a package is emerged, portage announces the start of the process, and the start of installation to the live disk. Well, now, it also announces completion of all work on the package. That causes another length of introductory string, so now there are three columns to scan for package numbers. It's a personal preference, I'm sure - even GOM(s) don't all agree. :) -- Regards, Peter.
Re: [gentoo-user] emerge -U or emerge -N
Neil Bothwick wrote: > On Tue, 11 Apr 2023 06:30:47 -0500, Dale wrote: > >>> -U: if a package doesn't need to be updated on your system today, >>> it'll be left until it does. >>> >>> -N: if any USE flag at all has changed in a package, it'll be >>> recompiled, whether it needs to be on your system or not. >>> >> >> When I say deeper, I mean it will find more packages that may not be >> found otherwise. > And -e finds even more - but more is not always better. -U was introduced > because -N was causing too many packages to be rebuilt unnecessarily. > > Before those options came along, I would run emerge -e world to fix problems. Sometimes revdep-rebuild would catch things but sometimes it wouldn't. Thing is, since I started using the current options, I have few problems with package upgrades. Sure, they have a known bug on occasion but recompiling won't help that. I'm just talking about problems with one package not matching up with some other package and recompile fixes it. To me, the fact it works so much better tells me I'm doing something right. Other people may get different results but as long as what I'm doing works, I don't plan to change anything. Dale :-) :-)
Re: [gentoo-user] emerge -U or emerge -N
On Tue, 11 Apr 2023 06:30:47 -0500, Dale wrote: > > -U: if a package doesn't need to be updated on your system today, > > it'll be left until it does. > > > > -N: if any USE flag at all has changed in a package, it'll be > > recompiled, whether it needs to be on your system or not. > > > > > When I say deeper, I mean it will find more packages that may not be > found otherwise. And -e finds even more - but more is not always better. -U was introduced because -N was causing too many packages to be rebuilt unnecessarily. -- Neil Bothwick "Logic will get you from A to B. Imagination will take you everywhere." (Albert Einstein) pgpqMA91uTQ4P.pgp Description: OpenPGP digital signature
Re: [gentoo-user] Portage 'Completed (m of n)' messages
On Tue, 11 Apr 2023 11:59:08 +0100, Peter Humphrey wrote: > What does the panel think of these new status messages from portage > (~amd64)? At first I thought "that's useful", but after a while I > concluded that it just adds to the clutter on the screen and actually > impedes my ability to follow progress. I like it. Previously, I had to wait for the completed counter to increment before knowing that installation of a package was complete. It's handy if you need to interrupt an emerge session and don't want to do it just before Chromium has finished installing :-O > Perhaps I'm just being a Grumpy Old Man (tm). See above, some Grumpy Old Men like it ;-) -- Neil Bothwick Some cause happiness wherever they go. Others whenever they go. pgp4HNzukR9Us.pgp Description: OpenPGP digital signature
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Tue, Apr 11, 2023 at 9:47 AM Matt Connell wrote: > > I usually try not to edit any files that are 'managed' by packages, but > sometimes it is unavoidable (eg. no thing.conf.d directory support), so > I wind up having to either accept the change and then re-edit it, or > zap the change and allow the file to get stale as the package is > updated, leading back to the first scenario regardless. > Yeah, conf.d is a much better paradigm, but cfg-update dates back to when that wasn't popular, and of course there are still plenty of packages that don't support it today. Plus when you do need to manually edit a file you can do it in meld with 3-way diffs. -- Rich
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Tue, 2023-04-11 at 09:34 -0400, Rich Freeman wrote: > cfg-update is a bit crufty, but its main advantage is support for > 3-way merges, which are usually automated. So if you change one line > in the middle of a config file you won't have to manually go through > diffs to re-apply the change every time it is updated. If the section > immediately around the line you edited didn't change, then it will > just accept the upstream changes while maintaining your customization. Neat. Thanks for the explanation. I might make use of cfg-update from time to time after all. I usually try not to edit any files that are 'managed' by packages, but sometimes it is unavoidable (eg. no thing.conf.d directory support), so I wind up having to either accept the change and then re-edit it, or zap the change and allow the file to get stale as the package is updated, leading back to the first scenario regardless.
Re: [gentoo-user] Re: Portage 'Completed (m of n)' messages
On Tue, 11 Apr 2023 at 14:45, Peter Humphrey wrote: > > On Tuesday, 11 April 2023 13:28:58 BST Nikos Chantziaras wrote: > > On 11/04/2023 13:59, Peter Humphrey wrote: > > > What does the panel think of these new status messages from portage > > > (~amd64)? > > What messages? Where? When? :P > > They seem to be a feature of sys-apps/portage-3.0.46 and probably 3.0.45. Nikos' questions still stand :P Cheers, Arve
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Tue, Apr 11, 2023 at 9:14 AM Matt Connell wrote: > > On Mon, 2023-04-10 at 23:44 -0600, the...@sys-concept.com wrote: > > After update I get: > > * IMPORTANT: config file '/etc/mtab' needs updating. > > > > What is this, don't remember seeing it before. > > > > cfg-update -u > > doesn't give me an option to view it. > > > > > > dispatch-conf will show you what is being changed and give you the > option to use/zap the change. > > I never even knew cfg-update existed (I've always used dispatch-conf). > I'm guessing cfg-update doesn't understand symlinks, and obviously they can't be conventionally edited. cfg-update is a bit crufty, but its main advantage is support for 3-way merges, which are usually automated. So if you change one line in the middle of a config file you won't have to manually go through diffs to re-apply the change every time it is updated. If the section immediately around the line you edited didn't change, then it will just accept the upstream changes while maintaining your customization. It uses RCS, which is obviously dated. I'm maintaining it, so patches are welcome, but I'm not really putting any effort into it. As others have pointed out, it should be a symlink to /proc/self/mounts, as with the increasingly more popular use of mount namespaces there is no system-wide concept of what is/isn't mounted where. With the symlink each process will see what is actually mounted in its own namespace, avoiding leaking info from the host namespace, and also avoiding providing information that is incorrect from the process's perspective. -- Rich
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Mon, 2023-04-10 at 23:44 -0600, the...@sys-concept.com wrote: > After update I get: > * IMPORTANT: config file '/etc/mtab' needs updating. > > What is this, don't remember seeing it before. > > cfg-update -u > doesn't give me an option to view it. > > dispatch-conf will show you what is being changed and give you the option to use/zap the change. I never even knew cfg-update existed (I've always used dispatch-conf).
Re: [gentoo-user] Re: Portage 'Completed (m of n)' messages
On Tuesday, 11 April 2023 13:28:58 BST Nikos Chantziaras wrote: > On 11/04/2023 13:59, Peter Humphrey wrote: > > What does the panel think of these new status messages from portage > > (~amd64)? > What messages? Where? When? :P They seem to be a feature of sys-apps/portage-3.0.46 and probably 3.0.45. -- Regards, Peter.
[gentoo-user] Re: config file '/etc/mtab' needs updating
On 11/04/2023 13:38, Dale wrote: I saw this the other day as well. I just skipped it. Still, it made me wonder, given what it does and what should update the file, why should emerge touch that file? Because the file belongs to the sys-apps/baselayout package. It's like fstab. I doubt I'd ever let any config update touch fstab. I don't think I'd ever try to touch mtab myself. fstab is a config file. mtab is not ;-) It doesn't even belong in /etc, but it's kept there for backwards compatibility.
[gentoo-user] Re: Portage 'Completed (m of n)' messages
On 11/04/2023 13:59, Peter Humphrey wrote: What does the panel think of these new status messages from portage (~amd64)? What messages? Where? When? :P
Re: [gentoo-user] emerge -U or emerge -N
Neil, On Tuesday, 2023-04-11 08:19:10 +0100, you wrote: > ... > So now we kn ow, ChatGPT is case-insensitive, it gave you answers for -u > and -n. You aren't really flabbergasted, are you? After all Microsoft is known for having a particularly soft spot for case-insensitiveness :-) Sincerely, Rainer
Re: [gentoo-user] emerge -U or emerge -N
Peter Humphrey wrote: > On Tuesday, 11 April 2023 11:33:38 BST Dale wrote: > >> The info from the man page is correct. > Of course it is. There'd be uproar if it weren't. > >> They do two different things. The -N will mean more recompiles of packages >> but it also means that when a USE flag change is made, it also changes any >> packages that relates to that. In other words, it goes deeper. > I don't know why you think it goes deeper, Dale. It's a question of candidate > selection, not depth of analysis. > > -U: if a package doesn't need to be updated on your system today, it'll be > left until it does. > > -N: if any USE flag at all has changed in a package, it'll be recompiled, > whether it needs to be on your system or not. > When I say deeper, I mean it will find more packages that may not be found otherwise. The deeper the hole I dig, the more dirt I have. That sort of thing. Maybe it should be phrased another way??? Dale :-) :-)
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Tuesday, 11 April 2023 12:03:57 BST Dr Rainer Woitok wrote: > Peter, > > On Tuesday, 2023-04-11 11:19:31 +0100, you wrote: > > On Tuesday, 11 April 2023 06:44:03 BST the...@sys-concept.com wrote: > > > After update I get: > > > * IMPORTANT: config file '/etc/mtab' needs updating. > > > ... > > > > /etc/mtab should be a symlink to /proc/self/mounts, so it ought to be kept > > up to date by the kernel as required. > > If I remember correctly, recently the symlink target was subtly changed > from "/proc/self/mounts" to "../proc/self/mounts", thus unconditionally > referring to a target on the same volume. Yes, I saw that, but is it even possible for /etc and /proc to be on different volumes? I thought not, but I'm always happy to be corrected. -- Regards, Peter.
Re: [gentoo-user] Jobs and load-average
On Wednesday, 15 February 2023 09:56:22 BST Peter Humphrey wrote: > Hello list, > > Not long ago I read that we should allow 2GB RAM for every emerge job - that > is, we should divide our RAM size by 2 to get the maximum number of > simultaneous jobs. I'm trying to get that right, but I'm not there yet. > > I have these entries in make.conf: > EMERGE_DEFAULT_OPTS="--jobs=16 --load-average=32 --autounmask=n --quiet- > unmerge-warn --ke> > MAKEOPTS="-j16" > > Today, though, I saw load averages going up to 72. Can anyone suggest better > values to suit my 24 threads and 64GB RAM? Thanks all for your contributions. I've settled on the following, after some experimenting: EMERGE_DEFAULT_OPTS="--jobs --autounmask=n --quiet-unmerge-warn --keep-going --nospinner" MAKEOPTS="-j24" I've stopped using disk space for /var/tmp/portage, even for the biggest packages, because (a) it causes a huge increase in compilation time, even on a SATA SSD, and (b) I've never seen an OOM anyway. So what if the system load goes high? It's only the number of processes ready for execution at any instant. I imagine the kernel is effective in guarding its own memory spaces. -- Regards, Peter.
Re: [gentoo-user] config file '/etc/mtab' needs updating
Peter, On Tuesday, 2023-04-11 11:19:31 +0100, you wrote: > On Tuesday, 11 April 2023 06:44:03 BST the...@sys-concept.com wrote: > > After update I get: > > * IMPORTANT: config file '/etc/mtab' needs updating. > > ... > /etc/mtab should be a symlink to /proc/self/mounts, so it ought to be kept up > to date by the kernel as required. If I remember correctly, recently the symlink target was subtly changed from "/proc/self/mounts" to "../proc/self/mounts", thus unconditionally referring to a target on the same volume. That's probably the reason for the warning thelma got. Sincerely, Rainer
[gentoo-user] Portage 'Completed (m of n)' messages
Hello list, What does the panel think of these new status messages from portage (~amd64)? At first I thought "that's useful", but after a while I concluded that it just adds to the clutter on the screen and actually impedes my ability to follow progress. Perhaps I'm just being a Grumpy Old Man (tm). -- Regards, Peter.
Re: [gentoo-user] emerge -U or emerge -N
On Tuesday, 11 April 2023 11:33:38 BST Dale wrote: > The info from the man page is correct. Of course it is. There'd be uproar if it weren't. > They do two different things. The -N will mean more recompiles of packages > but it also means that when a USE flag change is made, it also changes any > packages that relates to that. In other words, it goes deeper. I don't know why you think it goes deeper, Dale. It's a question of candidate selection, not depth of analysis. -U: if a package doesn't need to be updated on your system today, it'll be left until it does. -N: if any USE flag at all has changed in a package, it'll be recompiled, whether it needs to be on your system or not. -- Regards, Peter.
Re: [gentoo-user] Re: emerge -U or emerge -N
Neil Bothwick wrote: > On Tue, 11 Apr 2023 01:49:50 - (UTC), Grant Edwards wrote: > >>> I always do both except I use the lower case 'u'. I started using >>> Gentoo back in 2003. Over the years, I added/changed options to >>> emerge until I got a good sane system that works as expected and is >>> stable. My command is emerge -auDN world and it has worked for years. >>> >> Once upon a time, a little over 20 years ago, I did some studying, and >> I searched mailing lists postings for recommendations, and I settled >> on >> >> emerge -auvND > I used that for a while, except like Dale I prefer to not use -v (I like > the quiet life) but I switched to -U (--changed-use) to lower the number > of unnecessary rebuilds. > > What gets me, I have -v in make.conf for the default options. I shouldn't add it to the command line but I'm so used to doing so, I type it in anyway. It's a habit I just can't seem to break. :/ Dale :-) :-)
Re: [gentoo-user] config file '/etc/mtab' needs updating
the...@sys-concept.com wrote: > After update I get: > * IMPORTANT: config file '/etc/mtab' needs updating. > > What is this, don't remember seeing it before. > > cfg-update -u > doesn't give me an option to view it. > > > I saw this the other day as well. I just skipped it. Still, it made me wonder, given what it does and what should update the file, why should emerge touch that file? It's like fstab. I doubt I'd ever let any config update touch fstab. I don't think I'd ever try to touch mtab myself. I'm curious, is this a bug or is something touching something it shouldn't? Dale :-) :-)
Re: [gentoo-user] emerge -U or emerge -N
jul...@jroy.ca wrote: > On Mon, 2023-04-10 at 22:10 -0600, the...@sys-concept.com wrote: >> On 4/10/23 18:53, Dale wrote: >> >> >> I've asked ChatGPT for explanation and here is what I got: >> >> Here are the differences between emerge -U and emerge -N: >> >> emerge -U: This option upgrades the specified package(s) to the >> latest available version. It will first download the new version, >> then build and install it. If a dependency of the package being >> upgraded also needs to be upgraded, it will also be upgraded. >> >> emerge -N: This option installs the specified package(s) without >> upgrading any dependencies. It will only download and install the >> package(s) if they are not already installed. If any dependencies of >> the package(s) are not already installed, the command will fail. >> >> In other words, emerge -U upgrades packages and their dependencies, >> while emerge -N only installs packages without upgrading any >> dependencies. >> > This is a good example of why ChatGPT cannot be trusted. > When ChatGPT doesn't know the answer to something, rather than saying > it doesn't know the answer, it just makes it up. > > The difference between -U and -N as explained by ChatGPT is wrong; in > fact, it has nothing to do with dependencies. > > To have a truthful answer, let's not ask ChatGPT and instead look at > `man 5 emerge`: > > --newuse, -N > Tells emerge to include installed packages where USE > flags have changed since compilation. This option also > implies the --selective option. USE flag changes > include: > > A USE flag was added to a package. A USE flag was > removed from a package. A USE flag was > turned on for > a package. A USE flag was turned off for a package. > > --changed-use, -U > Tells emerge to include installed packages where USE > flags have changed since installation. This option > also implies the --selective option. Unlike --newuse, > the --changed-use option does not trigger > reinstallation when flags that the user has not > enabled are added orremoved. > > In a nutshell, `--newuse` or `-N` rebuilds packages when USE flags have > changed, regardless of whether the changed USE flags affect the outcome > Where as `--changed-use` or `-U` rebuilds packages when the USE flags > have changed, AND the changed USE flags affect the outcome. > > For example, suppose you are on an openRC system, and a package > introduces a new `systemd` USE flag; > With `-N`: this package will be rebuilt with `-systemd` > With `-U`: this package will not be rebuilt > The info from the man page is correct. They do two different things. The -N will mean more recompiles of packages but it also means that when a USE flag change is made, it also changes any packages that relates to that. In other words, it goes deeper. What all this comes down to, how stable and how consistent do you want your system to be? On some systems, it may get away with doing it the quick and fast way. In some cases it may not. It seemed to me that for my system, going a bit deeper worked better for me. I'd rather rebuild more packages and have a more stable system than take a quicker way and have problems every once in a while. If one wants to try the shorter way, see if it works for them, then that can be done. If it works, great. If not, switching to a method that takes longer and sorts through more packages may be needed. It's all up to the person sitting in the chair. Dale :-) :-)
Re: [gentoo-user] config file '/etc/mtab' needs updating
On Tuesday, 11 April 2023 06:44:03 BST the...@sys-concept.com wrote: > After update I get: > * IMPORTANT: config file '/etc/mtab' needs updating. > > What is this, don't remember seeing it before. > > cfg-update -u > doesn't give me an option to view it. /etc/mtab should be a symlink to /proc/self/mounts, so it ought to be kept up to date by the kernel as required. -- Regards, Peter.
Re: [gentoo-user] emerge -U or emerge -N
On Mon, 10 Apr 2023 22:10:32 -0600, the...@sys-concept.com wrote: > !'ve asked ChatGPT for explanation and here is what I got: > > Here are the differences between emerge -U and emerge -N: > > emerge -U: This option upgrades the specified package(s) to the latest > available version. It will first download the new version, then build > and install it. If a dependency of the package being upgraded also > needs to be upgraded, it will also be upgraded. > > emerge -N: This option installs the specified package(s) without > upgrading any dependencies. It will only download and install the > package(s) if they are not already installed. If any dependencies of > the package(s) are not already installed, the command will fail. So now we kn ow, ChatGPT is case-insensitive, it gave you answers for -u and -n. It's probably easier to read the man page than ask a bot to make a guess :( -- Neil Bothwick Hyperbole is absolutely the worst mistake you can possibly make pgppjQANgv7wq.pgp Description: OpenPGP digital signature
Re: [gentoo-user] Re: emerge -U or emerge -N
On Tue, 11 Apr 2023 01:49:50 - (UTC), Grant Edwards wrote: > > I always do both except I use the lower case 'u'. I started using > > Gentoo back in 2003. Over the years, I added/changed options to > > emerge until I got a good sane system that works as expected and is > > stable. My command is emerge -auDN world and it has worked for years. > > > > Once upon a time, a little over 20 years ago, I did some studying, and > I searched mailing lists postings for recommendations, and I settled > on > > emerge -auvND I used that for a while, except like Dale I prefer to not use -v (I like the quiet life) but I switched to -U (--changed-use) to lower the number of unnecessary rebuilds. -- Neil Bothwick Confucius say : He who play in root, eventually kill tree! pgpOcU12fvhNH.pgp Description: OpenPGP digital signature
