]> git.proxmox.com Git - pve-kernel-3.10.0.git/blob - README
bump version to 3.10.0-27
[pve-kernel-3.10.0.git] / README
1 KERNEL SOURCE:
2 ==============
3
4 We use the CENTOS7/RHEL7 kernel sources, available from:
5
6 http://vault.centos.org/
7
8
9 Additional/Updated Modules:
10 ---------------------------
11
12 - include latest ixgbe driver from intel/sourceforge
13
14 - include latest igb driver from intel/sourceforge
15
16 - include latest e1000e driver from intel/sourceforge
17
18 include latest Broadcom bnx2 drivers: http://driverdownloads.qlogic.com/
19
20 * original file linux-7.11.05.zip contains
21 netxtreme2-7.11.05.tar.gz (added to repository)
22
23 - include latest Adaptec aacraid driver
24
25 * http://download.adaptec.com/raid/aac/linux/aacraid-linux-src-1.2.1-40300.tgz
26
27 - include latest ARECA RAID drivers
28
29 * http://www.areca.com.tw/support/s_linux/linux.htm
30
31 - include native OpenZFS filesystem kernel modules for Linux
32
33 * https://github.com/zfsonlinux/
34
35 For licensing questions, see: http://open-zfs.org/wiki/Talk:FAQ
36
37 - include latest HPSA driver (HP Smart Array)
38
39 * http://sourceforge.net/projects/cciss/
40
41
42 FIRMWARE:
43 =========
44
45 We create our own firmware package, which includes the firmware for
46 all proxmox-ve kernels. So far this include
47
48 pve-kernel-2.6.18
49 pve-kernel-2.6.24
50 pve-kernel-2.6.32
51 pve-kernel-2.6.35
52 pve-kernel-3.10.0
53
54 We use 'find-firmware.pl' to extract lists of required firmeware
55 files. The script 'assemble-firmware.pl' is used to read those lists
56 and copy the files from various source directory into a target
57 directory.
58
59 We do not include firmeware for some wireless HW when there is a
60 separate debian package for that, for example:
61
62 zd1211-firmware
63 atmel-firmware
64 bluez-firmware
65
66
67 PATCHES:
68 --------
69
70 bridge-patch.diff: Avoid bridge problems with changing MAC
71 see also: http://forum.openvz.org/index.php?t=msg&th=5291
72
73 Behaviour after 2.6.27 has changed slighly - after setting mac address
74 of bridge device, then address won't change. So we could omit
75 that patch, requiring to set hwaddress in /etc/network/interfaces.
76
77
78 Additional information
79 ----------------------
80
81 We use the default configuration provided by RHEL, and apply
82 the following modification:
83
84 see config-3.10.0.diff
85
86 - enable CONFIG_CEPH_FS=m (request from user)
87
88 - enable common CONFIG_BLK_DEV_XXX to avoid hardware detection
89 problems (udev, undate-initramfs hav serious problems without that)
90
91 CONFIG_BLK_DEV_SD=y
92 CONFIG_BLK_DEV_SR=y
93 CONFIG_BLK_DEV_DM=y
94
95 - compile NBD and RBD modules
96 CONFIG_BLK_DEV_NBD=m
97 CONFIG_BLK_DEV_RBD=m
98
99 - set LOOP_MIN_COUNT to 8 (debian defaults)
100 CONFIG_BLK_DEV_LOOP_MIN_COUNT=8
101
102 - disable module signatures (CONFIG_MODULE_SIG)
103
104 - enable IBM JFS file system
105
106 This is disabled in RHEL kernel for no real reason, so we enable
107 it as requested by users (bug #64)
108
109 - enable apple HFS and HFSPLUS
110
111 This is disabled in RHEL kernel for no real reason, so we enable
112 it as requested by users
113
114 - enable CONFIG_BCACHE=m (requested by user)
115
116 - enable CONFIG_BRIDGE=y
117
118 Else we get warnings on boot, that
119 net.bridge.bridge-nf-call-iptables is an unknown key
120
121 - disable CONFIG_DEFAULT_SECURITY_SELINUX
122
123 Use same SELINUX config as debian kernels (CONFIG_DEFAULT_SECURITY_DAC=y)
124
125 - set CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y
126
127 because if not set, it can give some dynamic memory or cpu frequencies
128 change, and vms can crash (mainly windows guest).
129
130 see http://forum.proxmox.com/threads/18238-Windows-7-x64-VMs-crashing-randomly-during-process-termination?p=93273#post93273
131
132 - set CONFIG_CC_STACKPROTECTOR_REGULAR=y
133
134 gcc on debian wheezy does not supportCONFIG_CC_STACKPROTECTOR_STRONG
135
136 # Note: enable now for testing
137 #- disable CONFIG_BRIDGE_IGMP_SNOOPING
138 #
139 # To avoid problems with multicasting. Various users reported
140 # multipath problems, which can be avoided by turning off IGMP
141 # snooping.
142
143 # deadline is default anyways now
144 #- use 'deadline' as default scheduler
145 #
146 # This is the suggested setting for KVM. We also measure bad fsync
147 # performance with ext4 and cfq.
148
149 Testing final kernel with kvm
150 -----------------------------
151
152 kvm -kernel data/boot/vmlinuz-3.10.0-2-pve -initrd initrd.img-3.10.0-2-pve -append "vga=791 video=vesafb:ywrap,mtrr" /dev/zero
153