Re: filesystems and year 2038

2022-04-06 Thread Phillip Lougher
> On Tue, Apr 5, 2022 at 8:48 AM Dusty Mabe > What about squashfs? We use that for the live media, is that affected? No, Squashfs uses unsigned 32-bit ints, which will roll over in 2106. Phillip --- Squashfs author and maintainer. ___ devel mailing

Re: filesystems and year 2038

2022-04-06 Thread Neal Gompa
On Wed, Apr 6, 2022 at 5:52 PM Eric Sandeen wrote: > > On 4/4/22 2:51 PM, Justin Forbes wrote: > > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: > >> > >> Hi, creating a thread on this from: > >> https://github.com/coreos/fedora-coreos-config/pull/1650 > >> > >> Basically I'd propose that

Re: filesystems and year 2038

2022-04-06 Thread Eric Sandeen
On 4/4/22 2:51 PM, Justin Forbes wrote: > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: >> >> Hi, creating a thread on this from: >> https://github.com/coreos/fedora-coreos-config/pull/1650 >> >> Basically I'd propose that not just our default images have y2038-compatible >> filesystem

Re: filesystems and year 2038

2022-04-05 Thread Chris Murphy
On Tue, Apr 5, 2022 at 8:50 AM Chris Adams wrote: > > Once upon a time, Colin Walters said: > > Ah but with a 512M disk I do get 256 bit inodes, I bet that's the > > difference. > > It comes from /etc/mke2fs.conf... kind of. Below 512M, mke2fs chooses > to use the "small" config from there,

Re: filesystems and year 2038

2022-04-05 Thread Chris Murphy
On Tue, Apr 5, 2022 at 8:27 AM Colin Walters wrote: > Or, maybe it's an Anaconda thing to override it? I don't think so (from Workstation edition installation) # grep mke2 /var/log/anaconda/storage.log INFO:program:Running... mke2fs -t ext4 /dev/vda2 INFO:program:b'mke2fs 1.46.5 (30-Dec-2021)'

Re: filesystems and year 2038

2022-04-05 Thread Chris Adams
Once upon a time, Colin Walters said: > Ah but with a 512M disk I do get 256 bit inodes, I bet that's the difference. It comes from /etc/mke2fs.conf... kind of. Below 512M, mke2fs chooses to use the "small" config from there, which includes the smaller inode_size. The thresholds are hard-coded

Re: filesystems and year 2038

2022-04-05 Thread Neal Gompa
On Tue, Apr 5, 2022 at 10:27 AM Colin Walters wrote: > > > > On Tue, Apr 5, 2022, at 10:11 AM, Justin Forbes wrote: > > > > That list hasn't been edited in 5 years, but 256 bit inodes have been > > the ext default for a very long time unless you specifically request > > small. > > In current

Re: filesystems and year 2038

2022-04-05 Thread Colin Walters
On Tue, Apr 5, 2022, at 10:11 AM, Justin Forbes wrote: > > That list hasn't been edited in 5 years, but 256 bit inodes have been > the ext default for a very long time unless you specifically request > small. In current Fedora CoreOS we have 128 bit inodes for /boot, and this appears to be

Re: filesystems and year 2038

2022-04-05 Thread Richard W.M. Jones
On Mon, Apr 04, 2022 at 02:51:52PM -0500, Justin Forbes wrote: > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: > > > > Hi, creating a thread on this from: > > https://github.com/coreos/fedora-coreos-config/pull/1650 > > > > Basically I'd propose that not just our default images have > >

Re: filesystems and year 2038

2022-04-05 Thread Justin Forbes
On Tue, Apr 5, 2022 at 8:50 AM Neal Gompa wrote: > > On Tue, Apr 5, 2022 at 9:48 AM Colin Walters wrote: > > > > > > > > On Mon, Apr 4, 2022, at 3:51 PM, Justin Forbes wrote: > > > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: > > >> > > >> Hi, creating a thread on this from: > > >>

Re: filesystems and year 2038

2022-04-05 Thread Neal Gompa
On Tue, Apr 5, 2022 at 9:48 AM Colin Walters wrote: > > > > On Mon, Apr 4, 2022, at 3:51 PM, Justin Forbes wrote: > > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: > >> > >> Hi, creating a thread on this from: > >> https://github.com/coreos/fedora-coreos-config/pull/1650 > >> > >>

Re: filesystems and year 2038

2022-04-05 Thread Colin Walters
On Mon, Apr 4, 2022, at 3:51 PM, Justin Forbes wrote: > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: >> >> Hi, creating a thread on this from: >> https://github.com/coreos/fedora-coreos-config/pull/1650 >> >> Basically I'd propose that not just our default images have y2038-compatible

Re: filesystems and year 2038

2022-04-05 Thread Peter Robinson
On Tue, Apr 5, 2022 at 2:10 PM Chris Adams wrote: > > Once upon a time, Neal Gompa said: > > What about squashfs? We use that for the live media, is that affected? > > There's also vfat (for EFI system partition) and ISO9660 (base for all > media). How do they handle dates? Did you look at

Re: filesystems and year 2038

2022-04-05 Thread Justin Forbes
On Tue, Apr 5, 2022 at 8:10 AM Chris Adams wrote: > > Once upon a time, Neal Gompa said: > > What about squashfs? We use that for the live media, is that affected? > > There's also vfat (for EFI system partition) and ISO9660 (base for all > media). How do they handle dates? > -- FAT

Re: filesystems and year 2038

2022-04-05 Thread Chris Adams
Once upon a time, Neal Gompa said: > What about squashfs? We use that for the live media, is that affected? There's also vfat (for EFI system partition) and ISO9660 (base for all media). How do they handle dates? -- Chris Adams ___ devel mailing

Re: filesystems and year 2038

2022-04-05 Thread Neal Gompa
On Tue, Apr 5, 2022 at 8:48 AM Dusty Mabe wrote: > > > > On 4/5/22 03:54, Peter Robinson wrote: > > > > Is ext4 actually a problem here? > > It would appear not. I was going off the info from the original email and > just noticed that Justin's > reply addressed the xfs concern and I was

Re: filesystems and year 2038

2022-04-05 Thread Dusty Mabe
On 4/5/22 03:54, Peter Robinson wrote: > > Is ext4 actually a problem here? It would appear not. I was going off the info from the original email and just noticed that Justin's reply addressed the xfs concern and I was wondering about ext4. Thanks for the info. Dusty

Re: filesystems and year 2038

2022-04-05 Thread Kevin Kofler via devel
Peter Robinson wrote: > Is ext4 actually a problem here? From the y2038 list [1] the ext4 with > "new inodes" which has the option of "34 bit seconds / 30-bit ns" has > a rollover date of 2582. The new extended i_ctime_extra [2] was added > in 2007 and apparently made the default in 2008 [3] so it

Re: filesystems and year 2038

2022-04-05 Thread Peter Robinson
> >> Hi, creating a thread on this from: > >> https://github.com/coreos/fedora-coreos-config/pull/1650 > >> > >> Basically I'd propose that not just our default images have > >> y2038-compatible filesystem setups, we ensure that if e.g. XFS is > >> explicitly chosen for a Workstation

Re: filesystems and year 2038

2022-04-04 Thread Dusty Mabe
On 4/4/22 15:51, Justin Forbes wrote: > On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: >> >> Hi, creating a thread on this from: >> https://github.com/coreos/fedora-coreos-config/pull/1650 >> >> Basically I'd propose that not just our default images have y2038-compatible >> filesystem

Re: filesystems and year 2038

2022-04-04 Thread Justin Forbes
On Mon, Apr 4, 2022 at 11:47 AM Colin Walters wrote: > > Hi, creating a thread on this from: > https://github.com/coreos/fedora-coreos-config/pull/1650 > > Basically I'd propose that not just our default images have y2038-compatible > filesystem setups, we ensure that if e.g. XFS is explicitly

Re: filesystems and year 2038

2022-04-04 Thread Neal Gompa
On Mon, Apr 4, 2022 at 12:47 PM Colin Walters wrote: > > Hi, creating a thread on this from: > https://github.com/coreos/fedora-coreos-config/pull/1650 > > Basically I'd propose that not just our default images have y2038-compatible > filesystem setups, we ensure that if e.g. XFS is explicitly