On Mon, Mar 23, 2020 at 3:29 PM Edward Shishkin <edward.shishkin@xxxxxxxxx> wrote: > > On 03/23/2020 01:39 PM, Metztli Information Technology wrote: > > On Wed, Mar 11, 2020 at 3:09 PM Edward Shishkin <edward.shishkin@xxxxxxxxx> wrote: > >> > >> On 03/10/2020 11:24 PM, Edward Shishkin wrote: > >>> On 03/10/2020 09:25 PM, Edward Shishkin wrote: > >>>> On 03/10/2020 06:36 AM, Metztli Information Technology wrote: > >>>>> Niltze, Ed- > >>>>> > >>>>> > >>>>> I am experiencing segfault when using dpkg (package manager for > >>>>> debian) utility to unpack software. > >>>>> > >>>>> It happens *both* during reiser4 sfrn 5.1.3 -enabled Debian Installer > >>>>> (d-i) *and* in a pure sfrn 5.1.3 Debian Buster environment, > >>>>> i.e., formatted in sfrn 5.1.3 and running reiser4progs 2.0.0 and > >>>>> corresponding kernel. It happens in VirtualBox 6.1.4 > >>>> > >>>> Do you have any bootable image I could download and reproduce that > >>>> oops on my virtual machine? > >>> > >>> Ah, I see a mistake, which most likely causes that oops. > >>> Will provide a fixup soon.. > >> > >> > >> Please, apply the attached patch (from the directory ./fs/reiser4) > > > > Applied your patch, ed, and the reiser4 SFRN 5.1.3 native installation proceeds a little further --until after the kernel is selected-- the locales package fails. > > > > Here is a little background. The Debian Installer (d-i) utilizes Busybox utilities --especifically unxz-- once the 'Install base system' option starts. > > > > Notwithstanding, for whatever reason unxz fails with: > > > > ... > > unxz: compressed data not read from terminal, use -f to force it > > usage: /usr/lib/debootstrap/pkgdetails PKGS mirror packagesfile pkgs.. > > Â Â or: /usr/lib/debootstrap/pkgdetails FIELD field mirror packagesfile pkgs.. > > Â Â or: /usr/lib/debootstrap/pkgdetails GETDEPS packagesfile pkgs.. > > Â Â or: /usr/lib/debootstrap/pkgdetails STANZAS packagesfile pkgs.. > > Â Â or: /usr/lib/debootstrap/pkgdetails WGET% low high end reason > > mknod: /target/dev/null: No such file or directory > > > > I tried a couple (at least) of methods to bypass the issue. > > I located /usr/share/debootstrap/functions and modified Busybox command invocation as: unxz -f > > > > Additionally, I built another Busybox hack which included the full unxz command along with the proper *.so library support built from bleeding edge git source. > > < https://github.com/xz-mirror/xz > > > > > Fact is non-busybox unxz command does not require -f switch to continue the installation onto reiser4 SFRN5 formatted media: > > < https://metztli.it/buster-reiser5/reg42/metztli-reiser4-sfrn5.iso > > > < https://metztli.it/buster-reiser5/reg42/metztli-reiser4-sfrn5.iso.SHA256SUM > > > > > At this point is when the patch that you provided proved its worth --as the segfault was not experienced any longer and the installation continued. > > > > Notwithstanding, after selecting the kernel to be installed, for which the locales package is a requirement, the installer Python 'thinks' --like busybox unxz before-- > > that it is running in a terminal: > > sed: couldn't edit /etc/locale.gen: is a terminal > > > > Thinking that a newer Python3 version 3.8.2 might make a difference, I hacked a Debian UDEB component so as to provide Python3 3.8.2 support into > > the environment. Nevertheless, the issue persists, the Debian Installer continues to 'think' it is running in a terminal (read only? ) and not being installed > > into reiser4 SFRN5 formatted filesystem ( isatty() issues?) > > > > We can verify we are running non-busybox unxz utility: > > unxz --version > > > > and latest Python3 support: > > python3 --version > > > Hm.. I am not able to identify the problem only by that input.. > > > > > > Please be aware, sir, that even in a VirtualBox 6.1.4 virtual machine converted from reiser4 SFRN 4.0.2 to SFRN 5.1.3 by the cp -a back-and-forth method > > and subsequent fine tuning, i.e., /etc/fstab modification, /boot/grub/grub.cfg, reiser4progs 2.0.0 proper installation, etc., > > existing Python3 7.x.y from Debian repositories is used by the Debian package manager to install and upgrade software packages. > > Accordingly, that older Python3 version fails to detect that it is operating on a file system and, like the installer, 'thinks' it is in a terminal. > > Evidently, most software installation/upgrade operations fail with a python prompt: > >>>> > > > > Finally, I have hacked a couple of newer Debian package utilities with newer Python3 3.8.2 in an attempt to observe if something changes without any luck. > > apt_2.0.0+nmu1_amd64.deb > > apt-utils_2.0.0+nmu1_amd64.deb > > libapt-pkg6.0_2.0.0+nmu1_amd64.deb > > > > python3-apt_1.9.10+nmu1_amd64.deb > > python-apt_1.9.10+nmu1_amd64.deb > > python-apt-common_1.9.10+nmu1_all.deb > > > > In summary, a new busybox 1.31.1 unxz utility and Python3 7.x.y / 3.8.2 cannot detect the underlying reiser5 file system where they are to operate. > > I am able to replace unxz to advance the installation but I cannot make Python3 to recognize the reiser5 filesystem > > > There is no "reiser5 file system". At least, I didn't rename it. > "reiser5" stands for "reiser4 kernel module of SFRN 5.X.Y". I am using the term in your previous announcement, sir, < https://marc.info/?l=reiserfs-devel&m=157780043509663&w=2 > interchangeably. Notwithstanding, if you want me to be pedantic, let me rephrase the previous statement: "In summary, a new busybox 1.31.1 unxz utility and Python3 7.7.3 / 3.8.2 cannot detect the underlying reiser4 Software Framework Release Number (SFRN) 5.1.3 file system where they are to operate. I am able to replace unxz to advance the installation but I cannot make Python3 to recognize the reiser4 SFRN 5.1.3 filesystem." > > Thanks, > Edward. > > > Â on which it is to operate --as > > it erroneously 'believes' it is running in a terminal. > > > > By the way, the current Debian Installer (d-i) netboot framework for *both* reiser4 SFRN 4.0.2 and referenced SFRN 5.1.3 is very similar for the Linux 5.5.8 version. > > Yet the installer has no issues detecting a reiser4 4.0.2 -formatted media whereas unxz and python3 in reiser4 5.1.3 fail and possibly detect it as a terminal. > > > >> > >> Thanks, > >> Edward. > >> > >>> > >>>> > >>>> Ã? Â as well as in actual > >>>>> physical hardware, same BUG line. > >>>>> > >>>>> This BUG makes it impossible to carry out a native Debian Buster > >>>>> installation onto a default reiser4 SFRN 5.1.3 -formatted > >>>>> partition, i.e., reg42, with or without specifying a stripe, i.e., -t > >>>>> 256K. Accordingly, even if a user uses cp -a back-and-forth > >>>>> to hack a pure SFRN5 Debian system whenever a software upgrade occurs > >>>>> a segmentation fault will follow. > >>>>> > >>>>> Unaware if other package managers, like RPM/YUM utilities, etc., will > >>>>> cause a segfault as well. > >>>>> > >>>>> > >>>>> Mar Ã? 8 20:20:19 debootstrap: Selecting previously unselected package > >>>>> base-passwd. > >>>>> Mar Ã? 8 20:20:19 debootstrap: (Reading database ... 0 files and > >>>>> directories currently installed.) > >>>>> Mar Ã? 8 20:20:19 debootstrap: Preparing to unpack > >>>>> .../base-passwd_3.5.46_amd64.deb ... > >>>>> Mar Ã? 8 20:20:19 debootstrap: Segmentation fault > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378009] ------------[ cut here > >>>>> ]------------ > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378017] kernel BUG at > >>>>> fs/reiser4/plugin/item/extent_stripe_ops.c:466! > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378031] invalid opcode: 0000 [#1] SMP PTI > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378044] CPU: 6 PID: 18587 Comm: dpkg > >>>>> Tainted: G Ã? Â Ã? Â Ã? Â Ã? Â Ã? Â Ã? E Ã? Â Ã? Â 5.5.0-1+reiser4-5.1.3-amd64 #1 Debian > >>>>> 5.5.7-1+reiser4~5.1.3 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378048] Hardware name: Hewlett-Packard > >>>>> HP Pavilion dv6 Notebook PC/17FC, BIOS F.1C 01/23/2013 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378103] RIP: > >>>>> 0010:find_or_create_extent_stripe+0x169/0x1d0 [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378110] Code: 48 89 58 50 eb b3 4c 89 > >>>>> ef e8 43 40 fd ff eb 85 4c 89 e7 41 89 de e8 a6 2a d4 ee eb a2 4c 89 > >>>>> e7 e8 bc 3e d4 ee e9 11 ff ff ff <0f> 0b 48 89 df e8 bd 42 fb ff eb > >>>>> 89 e8 26 1e be ee f0 ff 4b 0c 74 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378114] RSP: 0018:ffff9f90cd873c30 > >>>>> EFLAGS: 00010246 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378120] RAX: 0000000000000000 RBX: > >>>>> ffff8ab298d12240 RCX: 0000000000000000 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378123] RDX: 0000000000000000 RSI: > >>>>> ffff8ab2a5b366f8 RDI: ffff8ab2a5b36600 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378127] RBP: ffff9f90cd873d20 R08: > >>>>> 0000000000000001 R09: 000000000000000c > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378130] R10: 0000000000000024 R11: > >>>>> ffff8ab29f350a82 R12: ffffd75c5061c1c0 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378134] R13: ffff8ab298d8a418 R14: > >>>>> 0000000000000000 R15: ffff8ab2b2905800 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378139] FS: Ã? 00007fc29b9f4380(0000) > >>>>> GS:ffff8ab2bfb80000(0000) knlGS:0000000000000000 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378143] CS: Ã? 0010 DS: 0000 ES: 0000 > >>>>> CR0: 0000000080050033 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378147] CR2: 00005616aa3b0a81 CR3: > >>>>> 00000004187d4002 CR4: 00000000000606e0 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378150] Call Trace: > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378211] Ã? ? shorten_stripe+0x39b/0x4d0 > >>>>> [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378221] Ã? ? _cond_resched+0x15/0x30 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378266] Ã? shorten_stripe+0x39b/0x4d0 > >>>>> [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378317] > >>>>> reiser4_setattr_generic+0x150/0x200 [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378362] Ã? ? shorten_stripe+0x4d0/0x4d0 > >>>>> [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378371] Ã? notify_change+0x339/0x4a0 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378378] Ã? do_truncate+0x74/0xc0 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378386] Ã? do_sys_ftruncate+0x130/0x180 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378395] Ã? do_syscall_64+0x52/0x170 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378401] > >>>>> entry_SYSCALL_64_after_hwframe+0x44/0xa9 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378406] RIP: 0033:0x7fc29bb81947 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378412] Code: 77 01 c3 48 8b 15 49 85 > >>>>> 0c 00 f7 d8 64 89 02 b8 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 > >>>>> 0f 1f 40 00 b8 4d 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 01 c3 48 8b > >>>>> 15 19 85 0c 00 f7 d8 64 89 02 b8 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378416] RSP: 002b:00007fff8148a888 > >>>>> EFLAGS: 00000206 ORIG_RAX: 000000000000004d > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378421] RAX: ffffffffffffffda RBX: > >>>>> 00005616aa3ad4f0 RCX: 00007fc29bb81947 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378425] RDX: 0000000000000000 RSI: > >>>>> 0000000000000098 RDI: 0000000000000009 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378428] RBP: 00005616aa3ac260 R08: > >>>>> 00005616aa3ac340 R09: 00007fc29b9f4380 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378432] R10: 0000000000000000 R11: > >>>>> 0000000000000206 R12: 0000000000000098 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378435] R13: 00007fff8148b2f0 R14: > >>>>> 00005616aa3b32a5 R15: 0000000000000000 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378440] Modules linked in: dm_mod(E) > >>>>> md_mod(E) xfs(E) reiser4(E) jfs(E) btrfs(E) xor(E) raid6_pq(E) > >>>>> libcrc32c(E) zstd_compress(E) zstd_decompress(E) ext4(E) crc16(E) > >>>>> mbcache(E) jbd2(E) crc32c_generic(E) r8169(E) realtek(E) libphy(E) > >>>>> ctr(E) ccm(E) iwldvm(E) mac80211(E) libarc4(E) iwlwifi(E) cfg80211(E) > >>>>> rfkill(E) nls_ascii(E) nls_cp437(E) vfat(E) fat(E) hid_generic(E) > >>>>> usbhid(E) hid(E) sr_mod(E) cdrom(E) sd_mod(E) uas(E) usb_storage(E) > >>>>> ahci(E) libahci(E) xhci_pci(E) ehci_pci(E) xhci_hcd(E) sdhci_pci(E) > >>>>> ehci_hcd(E) cqhci(E) libata(E) rtsx_pci_sdmmc(E) sdhci(E) usbcore(E) > >>>>> mmc_core(E) scsi_mod(E) rtsx_pci(E) mfd_core(E) usb_common(E) wmi(E) > >>>>> battery(E) video(E) [last unloaded: libphy] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378508] ---[ end trace > >>>>> c7be9d76279fc2ea ]--- > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378544] RIP: > >>>>> 0010:find_or_create_extent_stripe+0x169/0x1d0 [reiser4] > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378550] Code: 48 89 58 50 eb b3 4c 89 > >>>>> ef e8 43 40 fd ff eb 85 4c 89 e7 41 89 de e8 a6 2a d4 ee eb a2 4c 89 > >>>>> e7 e8 bc 3e d4 ee e9 11 ff ff ff <0f> 0b 48 89 df e8 bd 42 fb ff eb > >>>>> 89 e8 26 1e be ee f0 ff 4b 0c 74 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378554] RSP: 0018:ffff9f90cd873c30 > >>>>> EFLAGS: 00010246 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378558] RAX: 0000000000000000 RBX: > >>>>> ffff8ab298d12240 RCX: 0000000000000000 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378562] RDX: 0000000000000000 RSI: > >>>>> ffff8ab2a5b366f8 RDI: ffff8ab2a5b36600 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378565] RBP: ffff9f90cd873d20 R08: > >>>>> 0000000000000001 R09: 000000000000000c > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378568] R10: 0000000000000024 R11: > >>>>> ffff8ab29f350a82 R12: ffffd75c5061c1c0 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378572] R13: ffff8ab298d8a418 R14: > >>>>> 0000000000000000 R15: ffff8ab2b2905800 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378576] FS: Ã? 00007fc29b9f4380(0000) > >>>>> GS:ffff8ab2bfb80000(0000) knlGS:0000000000000000 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378580] CS: Ã? 0010 DS: 0000 ES: 0000 > >>>>> CR0: 0000000080050033 > >>>>> Mar Ã? 8 20:20:19 kernel: [ Ã? 794.378584] CR2: 00005616aa3b0a81 CR3: > >>>>> 00000004187d4002 CR4: 00000000000606e0 > >>>>> > >>>>> P.S. If, for the sake experiment, we format the target partition with reiserprogs 2.0.0 as mkfs.reiser4 -o compress=zstd1, we discover the fail installation experience is similar as if use mkfs.reiser4 -o create=reg42 ; could it be that the compression plugin, unsupported in SFRN 5.1.3, is somehow affecting the *overall* outcome. Is it possible to strip/disable the compression plugin from the reiser4 SFRN 5.1.3 code? Best Professional Regards. -- Jose R R http://metztli.it --------------------------------------------------------------------------------------------- Download Metztli Reiser4: Debian Buster w/ Linux 5.5.8 AMD64 --------------------------------------------------------------------------------------------- feats ZSTD compression https://sf.net/projects/metztli-reiser4/ ------------------------------------------------------------------------------------------- Official current Reiser4 resources: https://reiser4.wiki.kernel.org/
