From: Adhemerval Zanella <adhemerval.zanella@linaro.org>
To: libc-alpha@sourceware.org, Thorsten Kukuk <kukuk@suse.com>
Subject: [PATCH 0/3] Make accounting database no-op
Date: Thu, 9 Nov 2023 14:01:16 -0300 [thread overview]
Message-ID: <20231109170119.1664204-1-adhemerval.zanella@linaro.org> (raw)
The utmp/utmpx interface is the missing piece to enable full y2038
support on glibc, and even some 64 bit architectures are not fully
compatible (the ones that define __WORDSIZE_TIME64_COMPAT32).
There are multiple issues regarding 64 bit time_t support [1] [2] [3]
and a potential fix would either require breaking the ABI (as I
suggested when I started the y2038 and decided to put it aside due to
the required work) or use a different encoding for the timeval
structure (suggested by Andreas [4]).
The current implementation has some design flaws that are not
straightforward to fix without a complete rewrite [5] [6]. A utmp/utmpx
daemon will also require a security model and support for multiple
different IPC systems, which is out of scope for glibc. Also, this is
what systemd-logind essentially does so it would be a double effort
(the pt_chwon daemon is a remind that it is not easy to get this right).
This patchset removes all accounting database implementation and makes
the function no-op and/or return an error. There is not much gain in
moving the current implementation to compat symbols, it does not solve
the 64 bit time_t support for old binaries, nor it is guaranteed that
the UTMP/UTMPX files will exist in future environments.
Keeping a compat symbol also does not help with some design flags
like BZ#24492, which I am not sure why it did not raise more security
concerns since it is easy to create DoS attacks by preventing utmp
updates.
The utmp.h/utmpx.h headers are kept as is, even though glibc does not
use its definition. The related path _PATH_UTMP/_PATH_WTMP/etc. points
to invalid paths, and stub link warnings are added to the affected
symbols.
[1] https://sourceware.org/bugzilla/show_bug.cgi?id=28146
[2] https://sourceware.org/bugzilla/show_bug.cgi?id=17470
[3] https://sourceware.org/bugzilla/show_bug.cgi?id=30701
[4] https://sourceware.org/pipermail/libc-alpha/2023-August/150661.html
Adhemerval Zanella (3):
sysdeps: Re-flow and sort multiline gnu/Makefile definitions
login: Remove utmp fallback for getlogin
login: Make user accounting database no-op
NEWS | 18 +
include/set-freeres.h | 3 -
include/unistd.h | 3 -
include/utmp.h | 29 -
login/Makefile | 9 +-
login/endutxent.c | 25 -
login/getlogin.c | 13 +-
login/getutent.c | 28 +-
login/getutent_r.c | 45 +-
login/getutid.c | 25 +-
login/getutid_r.c | 34 +-
login/getutline.c | 27 +-
login/getutline_r.c | 21 +-
login/getutmp.c | 12 +-
login/getutmpx.c | 34 --
login/getutxent.c | 25 -
login/getutxid.c | 25 -
login/getutxline.c | 25 -
login/login.c | 118 +---
login/logout.c | 44 +-
login/logwtmp.c | 22 +-
login/programs/utmpdump.c | 62 ---
login/pututxline.c | 25 -
login/setutxent.c | 25 -
login/tst-pututxline-cache.c | 193 -------
login/tst-pututxline-lockfail.c | 176 ------
login/tst-updwtmpx.c | 112 ----
login/tst-utmp.c | 377 -------------
login/tst-utmpx.c | 2 -
login/updwtmp.c | 13 +-
login/updwtmpx.c | 25 -
login/utmp-private.h | 44 --
login/utmp_file.c | 506 ------------------
login/utmpname.c | 57 +-
login/utmpxname.c | 25 -
malloc/set-freeres.c | 6 -
manual/users.texi | 446 ++-------------
sysdeps/generic/paths.h | 6 +-
sysdeps/gnu/Makefile | 43 +-
sysdeps/gnu/getutmp.c | 34 --
sysdeps/gnu/getutmpx.c | 1 -
sysdeps/gnu/updwtmp.c | 30 --
sysdeps/gnu/utmp_file.c | 30 --
sysdeps/mach/hurd/getlogin.c | 35 --
sysdeps/unix/getlogin.c | 81 ---
sysdeps/unix/getlogin_r.c | 103 ----
sysdeps/unix/sysv/linux/getlogin.c | 39 --
sysdeps/unix/sysv/linux/getlogin_r.c | 34 +-
sysdeps/unix/sysv/linux/paths.h | 4 +-
sysdeps/unix/sysv/linux/s390/s390-32/Makefile | 4 -
.../unix/sysv/linux/s390/s390-32/getutent.c | 19 +-
.../unix/sysv/linux/s390/s390-32/getutent_r.c | 11 +-
.../unix/sysv/linux/s390/s390-32/getutid.c | 4 +-
.../unix/sysv/linux/s390/s390-32/getutid_r.c | 2 +-
.../unix/sysv/linux/s390/s390-32/getutline.c | 4 +-
.../sysv/linux/s390/s390-32/getutline_r.c | 2 +-
.../unix/sysv/linux/s390/s390-32/getutmp.c | 20 +-
.../unix/sysv/linux/s390/s390-32/getutxent.c | 29 -
.../unix/sysv/linux/s390/s390-32/getutxid.c | 29 -
.../unix/sysv/linux/s390/s390-32/getutxline.c | 29 -
sysdeps/unix/sysv/linux/s390/s390-32/login.c | 1 +
.../unix/sysv/linux/s390/s390-32/login32.c | 37 --
.../unix/sysv/linux/s390/s390-32/pututxline.c | 29 -
.../unix/sysv/linux/s390/s390-32/updwtmp.c | 6 +-
.../unix/sysv/linux/s390/s390-32/updwtmpx.c | 29 -
.../sysv/linux/s390/s390-32/utmp-convert.h | 85 ---
sysdeps/unix/sysv/linux/s390/s390-32/utmp32.c | 183 -------
sysdeps/unix/sysv/linux/s390/s390-32/utmp32.h | 51 --
.../sysv/linux/s390/s390-32/utmpx-convert.h | 84 ---
.../unix/sysv/linux/s390/s390-32/utmpx32.c | 138 -----
.../unix/sysv/linux/s390/s390-32/utmpx32.h | 59 --
sysdeps/unix/sysv/linux/utmp_file.c | 36 --
72 files changed, 212 insertions(+), 3798 deletions(-)
delete mode 100644 login/endutxent.c
delete mode 100644 login/getutmpx.c
delete mode 100644 login/getutxent.c
delete mode 100644 login/getutxid.c
delete mode 100644 login/getutxline.c
delete mode 100644 login/programs/utmpdump.c
delete mode 100644 login/pututxline.c
delete mode 100644 login/setutxent.c
delete mode 100644 login/tst-pututxline-cache.c
delete mode 100644 login/tst-pututxline-lockfail.c
delete mode 100644 login/tst-updwtmpx.c
delete mode 100644 login/tst-utmp.c
delete mode 100644 login/tst-utmpx.c
delete mode 100644 login/updwtmpx.c
delete mode 100644 login/utmp-private.h
delete mode 100644 login/utmp_file.c
delete mode 100644 login/utmpxname.c
delete mode 100644 sysdeps/gnu/getutmp.c
delete mode 100644 sysdeps/gnu/getutmpx.c
delete mode 100644 sysdeps/gnu/updwtmp.c
delete mode 100644 sysdeps/gnu/utmp_file.c
delete mode 100644 sysdeps/mach/hurd/getlogin.c
delete mode 100644 sysdeps/unix/getlogin.c
delete mode 100644 sysdeps/unix/getlogin_r.c
delete mode 100644 sysdeps/unix/sysv/linux/getlogin.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/getutxent.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/getutxid.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/getutxline.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/login32.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/pututxline.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/updwtmpx.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmp-convert.h
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmp32.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmp32.h
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmpx-convert.h
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmpx32.c
delete mode 100644 sysdeps/unix/sysv/linux/s390/s390-32/utmpx32.h
delete mode 100644 sysdeps/unix/sysv/linux/utmp_file.c
--
2.34.1
next reply other threads:[~2023-11-09 17:01 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-09 17:01 Adhemerval Zanella [this message]
2023-11-09 17:01 ` [PATCH 1/3] sysdeps: Re-flow and sort multiline gnu/Makefile definitions Adhemerval Zanella
2023-11-09 17:01 ` [PATCH 2/3] login: Remove utmp fallback for getlogin Adhemerval Zanella
2023-11-09 17:01 ` [PATCH 3/3] login: Make user accounting database no-op Adhemerval Zanella
2023-11-09 19:49 ` Joseph Myers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20231109170119.1664204-1-adhemerval.zanella@linaro.org \
--to=adhemerval.zanella@linaro.org \
--cc=kukuk@suse.com \
--cc=libc-alpha@sourceware.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).