diff options
author | Michael Kerrisk <mtk.manpages@gmail.com> | 2014-09-21 12:37:03 +0200 |
---|---|---|
committer | Michael Kerrisk <mtk.manpages@gmail.com> | 2014-09-21 12:37:03 +0200 |
commit | 95cef60aca7bfc35e4ad858ef642ee3ce82331b0 (patch) | |
tree | 84856917c776d6b91ef5336d3f0ba13b8211cb03 /Changes.old | |
parent | a2bac51d98775f423951b24cbce05317f0c5e864 (diff) |
Start of man-pages-3.74: updating Changes and Changes.old
Diffstat (limited to 'Changes.old')
-rw-r--r-- | Changes.old | 173 |
1 files changed, 173 insertions, 0 deletions
diff --git a/Changes.old b/Changes.old index 997948d91..3c9180ba1 100644 --- a/Changes.old +++ b/Changes.old @@ -33640,3 +33640,176 @@ packet.7 It looks like most of the socket options from this man pages are not defined in <netpacket/packet.h>. They are defined in <linux/if_packet.h> so we should include that one. + + +==================== Changes in man-pages-3.73 ==================== + +Released: 2014-09-21, Munich + + +Contributors +------------ + +The following people contributed patches/fixes or (noted in brackets +in the changelog below) reports, notes, and ideas that have been +incorporated in changes in this release: + +David Prévot <taffit@debian.org> +Eric W. Biederman <ebiederm@xmission.com> +J. Bruce Fields <bfields@fieldses.org> +Justin Cormack <justin@street-vision.com> +Lorenzo Beretta <lory.fulgi@infinito.it> +Michael Kerrisk <mtk.manpages@gmail.com> +Rob Landley <rob@landley.net> +Serge E. Hallyn <serge@hallyn.com> +Serge Hallyn <serge.hallyn@ubuntu.com> +Vasily Kulikov <segoon@openwall.com> +Vincent Lefevre <vincent@vinc17.net> +Vitaly Rybnikov <frodox@zoho.com> +Yuri Kozlov <yuray@komyakino.ru> + +Apologies if I missed anyone! + + +New and rewritten pages +----------------------- + +namespaces.7 + Michael Kerrisk [Eric W. Biederman] + New page providing overview of Linux namespaces + +pid_namespaces.7 + Michael Kerrisk [Eric W. Biederman, Vasily Kulikov, Rob Landley, + Serge Hallyn] + New page describing PID namesapces + +user_namespaces.7 + Michael Kerrisk [Eric W. Biederman, Andy Lutomirski, Serge Hallyn] + New page describing user namespaces. + + +Newly documented interfaces in existing pages +--------------------------------------------- + +clone.2 + Eric W. Biederman [Michael Kerrisk] + Document CLONE_NEWUSER for creating a new user namespace + +setns.2 + Eric W. Biederman, Michael Kerrisk + Document the PID, user, and mount namespace support + Document CLONE_NEWPID, CLONE_NEWUSER, and CLONE_NEWNS flags. + +unshare.2 + Michael Kerrisk [Eric W. Biederman] + Document CLONE_NEWPID + Michael Kerrisk [Eric W. Biederman] + Document CLONE_NEWUSER + Michael Kerrisk + Document CLONE_THREAD, CLONE_SIGHAND, and CLONE_VM + + +Changes to individual pages +--------------------------- + +clone.2 + Michael Kerrisk + Move some CLONE_NEWNET text to namespaces.7 + Michael Kerrisk + Move some CLONE_NEWUTS text 2 to namespaces.7 + Michael Kerrisk + Move some CLONE_NEWIPC text to namespaces.7 + Michael Kerrisk + Reword discussion of CLONE_NEWNS, removing text also in namespaces(7) + Michael Kerrisk + Standardize text on CLONE_NEW* flags and CAP_SYS_ADMIN + Michael Kerrisk + EINVAL if (CLONE_NEWUSER|CLONE_NEWPID) && (CLONE_THREAD|CLONE_PARENT) + Michael Kerrisk + Add more detail on the meaning of CLONE_SYSVSEM + +flock.2 + Michael Kerrisk [J. Bruce Fields] + Don't mention "classical BSD" in discussion of fcntl()/flock interaction + The noninteraction of flock(2) and fcntl(2) locks does + not seem to be classical BSD semantics (at least, checking + the 4.4BSD sources suggest that the lock types do interact, + although there have been other systems also where fcntl() + and flock() locks do not interact). So, fix the text + discussing "classical BSD" lock semantics. + +getunwind.2 + Michael Kerrisk [Yuri Kozlov] + Fix description of return value + s/size of unwind table/size of the unwind data/ + +mount.2 + Eric W. Biederman + Clarify use of 'mountflags' and 'data' for MS_REMOUNT + +reboot.2 + Michael Kerrisk [Justin Cormack, Eric W. Biederman] + Document effect of reboot() inside PID namespaces + +semop.2 + Michael Kerrisk + Refer to clone(2) for semantics of CLONE_SYSVSEM and semadj lists + +seteuid.2 +setgid.2 +setresuid.2 +setreuid.2 +setuid.2 + Michael Kerrisk + EINVAL can occur if UID/GID is not valid in caller's user namespace + +setns.2 + Michael Kerrisk [Eric W. Biederman] + Clarify capabilities required for reassociating with a mount namespace + Michael Kerrisk + Specify kernel version on each CLONE_NEW* flag + And remove text on flags from VERSIONS. + +unshare.2 + Michael Kerrisk + Add an example program + Michael Kerrisk + Clarify semantics of CLONE_SYSVSEM + Michael Kerrisk + CLONE_SYSVSEM does not require CAP_SYS_ADMIN + Michael Kerrisk + Note flags implied by CLONE_THREAD and CLONE_VM + +clock.3 + Michael Kerrisk [Vincent Lefevre] + The implementation uses clock_gettime() was to improve *accuracy* + (The man page text mistakenly used the word "precision".) + +drand48.3 + Michael Kerrisk [Lorenzo Beretta] + Remove crufty text about SVID 3 marking drand48() obsolete + See http://bugs.debian.org/758293 + +proc.5 + Michael Kerrisk + Move /proc/[pid]/mounts text to namespaces.7 + Michael Kerrisk + Move /proc/[pid]/mountstats text to namespaces.7 + +capabilities.7 + Michael Kerrisk + Refer reader to user_namespaces(7) for a discussion of capabilities + Michael Kerrisk + Document CAP_SETUID and CAP_SETGID for user namespace mappings + Michael Kerrisk + setns() needs CAP_SYS_ADMIN in the *target* namespace + Michael Kerrisk + Since Linux 3.8, user namespaces no longer require CAP_SYS_ADMIN + +mq_overview.7 + Michael Kerrisk + Refer to namespaces(7) for info on POSIX MQs and IPC namespaces + +svipc.7 + Michael Kerrisk + Refer to namespaces(7) for info on System V IPC and IPC namespaces |