[pve-devel] [PATCH vncterm 1/5] bump DH_BITS to 2048

2017-02-02 Thread Fabian Grünbichler
Signed-off-by: Fabian Grünbichler --- this only affects Jessie, see next patch vncterm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vncterm.c b/vncterm.c index a077ea3..0e33791 100644 --- a/vncterm.c +++ b/vncterm.c @@ -169,7 +169,7 @@ static void vnc_debug_gnutls_log(in

[pve-devel] [PATCH vncterm 0/5] vncterm changes

2017-02-02 Thread Fabian Grünbichler
various build and gnutls related changes. most important are the changed cipher suites, I tested them with tigervnc and they seem to work - but more testing would probably be good. note that they are only used for the legacy VNC, not our builtin webconsole with noVNC. Fabian Grünbichler (5): bu

[pve-devel] [PATCH vncterm 2/5] use builtin dh_params with newer gnutls

2017-02-02 Thread Fabian Grünbichler
this is recommended instead of generating our own, but needs a new enough gnutls version. Signed-off-by: Fabian Grünbichler --- vncterm.c | 11 ++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/vncterm.c b/vncterm.c index 0e33791..0614e4b 100644 --- a/vncterm.c +++ b/vnct

[pve-devel] [PATCH vncterm 4/5] add dependency for new libgnutls30

2017-02-02 Thread Fabian Grünbichler
this allows building on Debian Jessie and Stretch Signed-off-by: Fabian Grünbichler --- debian/control | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/debian/control b/debian/control index e5f2e29..b130af7 100644 --- a/debian/control +++ b/debian/control @@ -2,7 +2,7 @@ Packa

[pve-devel] [PATCH vncterm 3/5] use builtin gnutls' builtin NORMAL ciphers

2017-02-02 Thread Fabian Grünbichler
instead of custom, outdated, insecure ones. NORMAL provides a sane default. note that this is only used for the legacy VNC anyway, since the noVNC web console (re-)uses the HTTPS connection to the API. Signed-off-by: Fabian Grünbichler --- vncterm.c | 40

[pve-devel] [PATCH vncterm 5/5] build: use temp file for glyph.h

2017-02-02 Thread Fabian Grünbichler
otherwise it gets truncated on failure, and make thinks it's there on the next run. Signed-off-by: Fabian Grünbichler --- Makefile | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/Makefile b/Makefile index abb85a6..c5523fb 100644 --- a/Makefile +++ b/Makefile @@ -23,7 +23

[pve-devel] applied: [PATCH kvm 2/2] various fixes

2017-02-02 Thread Wolfgang Bumiller
--- ...eck-transfer-mode-register-in-multi-block.patch | 61 ++ ...ock-count-enable-not-relevant-in-single-b.patch | 42 +++ ...tio-gpu-3d-check-virgl-capabilities-max_s.patch | 44 ...-10155-watchdog-6300esb-add-exit-function.patch | 50 +++

[pve-devel] applied: [PATCH kvm 1/2] fix #1237: include cirrus follow up fixes

2017-02-02 Thread Wolfgang Bumiller
This fixes CVE-2017-2615 together with pattern-fill regressions in the ciruss driver. introduced by the fixes for CVE-2016-9921 and CVE-2016-9922. --- For the record, cirrus is still the default display in many cases for us, but that is about to change since it's actually also deprecated. ...le-n

[pve-devel] [PATCH lxc+container] fix pvedaemon restart issues

2017-02-02 Thread Wolfgang Bumiller
As explained in the pve-container patch's cover letter: Until now the lxc-start process was sort of a danglign process in the pvedaemon.service cgroup when a container was started from the web UI causing long stalls when trying to restart pvedaemon and potential container kills. (Mostly

[pve-devel] [PATCH lxc] use forking mode for lxc@.service

2017-02-02 Thread Wolfgang Bumiller
--- debian/patches/series| 1 + debian/patches/use-forking-systemd-service.patch | 34 2 files changed, 35 insertions(+) create mode 100644 debian/patches/use-forking-systemd-service.patch diff --git a/debian/patches/series b/debian/patches/s

[pve-devel] [PATCH container] use the lxc@ service file provded by lxc

2017-02-02 Thread Wolfgang Bumiller
Until now the lxc-start process was sort of a danglign process in the pvedaemon.service cgroup when a container was started from the web UI causing long stalls when trying to restart pvedaemon and potential container kills. (Mostly problematic when issuing package upgrades). We now start containers

[pve-devel] [PATCH vncterm 2/4] libvncserver: refresh TLS patch

2017-02-02 Thread Fabian Grünbichler
Signed-off-by: Fabian Grünbichler --- vncpatches/tls-auth-pluging.patch | 45 +++ 1 file changed, 22 insertions(+), 23 deletions(-) diff --git a/vncpatches/tls-auth-pluging.patch b/vncpatches/tls-auth-pluging.patch index 821ffe1..17a8c47 100644 --- a/vncpatch

[pve-devel] [PATCH vncterm 4/4] build: update options for libvncserver 0.9.11

2017-02-02 Thread Fabian Grünbichler
Signed-off-by: Fabian Grünbichler --- Makefile | 11 ++- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/Makefile b/Makefile index c5523fb..6318f57 100644 --- a/Makefile +++ b/Makefile @@ -7,9 +7,10 @@ PACKAGERELEASE=1 ARCH:=$(shell dpkg-architecture -qDEB_BUILD_ARCH) CDAT

[pve-devel] [PATCH vncterm 3/4] libvncserver: add revert of systemd socket activation

2017-02-02 Thread Fabian Grünbichler
Signed-off-by: Fabian Grünbichler --- vncpatches/remove-systemd-socket-activation.patch | 227 ++ vncpatches/series | 1 + 2 files changed, 228 insertions(+) create mode 100644 vncpatches/remove-systemd-socket-activation.patch diff --git a/v

[pve-devel] [PATCH vncterm 1/4] libvncserver: update sources to 0.9.11

2017-02-02 Thread Fabian Grünbichler
Signed-off-by: Fabian Grünbichler --- SHA256: 193d630372722a532136fd25c5326b2ca1a636cbb8bf9bb115ef869c804d2894 URL: https://github.com/LibVNC/libvncserver/archive/LibVNCServer-0.9.11.tar.gz LibVNCServer-0.9.11.tar.gz | Bin 0 -> 1413739 bytes LibVNCServer-0.9.7.tar.gz | Bin 1105538 -> 0 bytes

[pve-devel] [PATCH vncterm 0/4] update libvncserver to 0.9.11

2017-02-02 Thread Fabian Grünbichler
this series updates libvncserver to the current release. I did some rudimentary testing with tigervnc and novnc, and notice neither breakage nor slowdowns. this is based on the previous gnutls patch series. (binary diff not included) Fabian Grünbichler (4): libvncserver: update sources to 0.9

Re: [pve-devel] [PATCH lxc] use forking mode for lxc@.service

2017-02-02 Thread Fabian Grünbichler
small comment inline On Thu, Feb 02, 2017 at 11:21:39AM +0100, Wolfgang Bumiller wrote: > --- > debian/patches/series| 1 + > debian/patches/use-forking-systemd-service.patch | 34 > > 2 files changed, 35 insertions(+) > create mode 100644 d

Re: [pve-devel] [PATCH lxc] use forking mode for lxc@.service

2017-02-02 Thread Wolfgang Bumiller
On Thu, Feb 02, 2017 at 12:12:33PM +0100, Fabian Grünbichler wrote: > > + Delegate=yes > > Shouldn't this last line be patched out as well if we are now using this > service? At least for Jessie's systemd, "Delegate" is not yet supported > and triggers one warning on start and two on stop ;) Yes.

[pve-devel] applied: [PATCH lxc] use forking mode for lxc@.service

2017-02-02 Thread Wolfgang Bumiller
applied, with the Delegate option change in a follow up patch On Thu, Feb 02, 2017 at 11:21:39AM +0100, Wolfgang Bumiller wrote: > --- > debian/patches/series| 1 + > debian/patches/use-forking-systemd-service.patch | 34 > > 2 files changed,

[pve-devel] applied: [PATCH container] use the lxc@ service file provded by lxc

2017-02-02 Thread Wolfgang Bumiller
On Thu, Feb 02, 2017 at 11:21:38AM +0100, Wolfgang Bumiller wrote: > Until now the lxc-start process was sort of a danglign > process in the pvedaemon.service cgroup when a container was > started from the web UI causing long stalls when trying to > restart pvedaemon and potential container kills.

[pve-devel] [PATCH guest-common v2 1/2] AbstractMigrate: remove unused rsync and scp remainders

2017-02-02 Thread Thomas Lamprecht
We have no call using rsync_cmd or scp_cmd here or any child class. The only part where rsync is used is in the PVE::Storage::storage_migrate call, but here both the command and the RSYNC_RSH environment variable gets assembled there locally, so clean this up here. Signed-off-by: Thomas Lamprecht

[pve-devel] [PATCH cluster v2] add ssh command helpers

2017-02-02 Thread Thomas Lamprecht
Add two helpers regarding ssh commands: * get_ssh_base_cmd: returns a ssh command with various standard options set, useful if only the base ssh command is needed * get_ssh_cmd: returns an string array for executing a command in a safe way on another pve node Both function ensure that we use

[pve-devel] [PATCH guest-common v2 2/2] AbstractMigrate: use get_ssh_base_cmd to assemble ssh_cmd

2017-02-02 Thread Thomas Lamprecht
this sets the HostKeyAlias for us, so that users using the dedicated migration network do not need to add each cluster node to every cluster members known_hosts file for each network they want to use Signed-off-by: Thomas Lamprecht --- changes since v1: * rebased on new guest-common package PV

[pve-devel] LXC with 2 and more NICs

2017-02-02 Thread Detlef Bracker
Dear, I thing so, thats a bug! A ping from outside to the LXC-containers to all NICs works fine! A ping from console via the NICs 2- is not possible! So, this can been a big problem, when a daemon will send from the NICs 2- ping 8.8.8.8 -I eth0 works fine ping 8.8.8.8 -I eth1 Destinati

[pve-devel] applied: [PATCH vncterm 0/5] vncterm changes

2017-02-02 Thread Dietmar Maurer
applied ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel