ia64/linux-2.6.18-xen.hg

annotate Documentation/networking/net-modules.txt @ 897:329ea0ccb344

balloon: try harder to balloon up under memory pressure.

Currently if the balloon driver is unable to increase the guest's
reservation it assumes the failure was due to reaching its full
allocation, gives up on the ballooning operation and records the limit
it reached as the "hard limit". The driver will not try again until
the target is set again (even to the same value).

However it is possible that ballooning has in fact failed due to
memory pressure in the host and therefore it is desirable to keep
attempting to reach the target in case memory becomes available. The
most likely scenario is that some guests are ballooning down while
others are ballooning up and therefore there is temporary memory
pressure while things stabilise. You would not expect a well behaved
toolstack to ask a domain to balloon to more than its allocation nor
would you expect it to deliberately over-commit memory by setting
balloon targets which exceed the total host memory.

This patch drops the concept of a hard limit and causes the balloon
driver to retry increasing the reservation on a timer in the same
manner as when decreasing the reservation.

Also if we partially succeed in increasing the reservation
(i.e. receive less pages than we asked for) then we may as well keep
those pages rather than returning them to Xen.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
author Keir Fraser <keir.fraser@citrix.com>
date Fri Jun 05 14:01:20 2009 +0100 (2009-06-05)
parents 831230e53067
children
rev   line source
ian@0 1 Wed 2-Aug-95 <matti.aarnio@utu.fi>
ian@0 2
ian@0 3 Linux network driver modules
ian@0 4
ian@0 5 Do not mistake this for "README.modules" at the top-level
ian@0 6 directory! That document tells about modules in general, while
ian@0 7 this one tells only about network device driver modules.
ian@0 8
ian@0 9 This is a potpourri of INSMOD-time(*) configuration options
ian@0 10 (if such exists) and their default values of various modules
ian@0 11 in the Linux network drivers collection.
ian@0 12
ian@0 13 Some modules have also hidden (= non-documented) tunable values.
ian@0 14 The choice of not documenting them is based on general belief, that
ian@0 15 the less the user needs to know, the better. (There are things that
ian@0 16 driver developers can use, others should not confuse themselves.)
ian@0 17
ian@0 18 In many cases it is highly preferred that insmod:ing is done
ian@0 19 ONLY with defining an explicit address for the card, AND BY
ian@0 20 NOT USING AUTO-PROBING!
ian@0 21
ian@0 22 Now most cards have some explicitly defined base address that they
ian@0 23 are compiled with (to avoid auto-probing, among other things).
ian@0 24 If that compiled value does not match your actual configuration,
ian@0 25 do use the "io=0xXXX" -parameter for the insmod, and give there
ian@0 26 a value matching your environment.
ian@0 27
ian@0 28 If you are adventurous, you can ask the driver to autoprobe
ian@0 29 by using the "io=0" parameter, however it is a potentially dangerous
ian@0 30 thing to do in a live system. (If you don't know where the
ian@0 31 card is located, you can try autoprobing, and after possible
ian@0 32 crash recovery, insmod with proper IO-address..)
ian@0 33
ian@0 34 --------------------------
ian@0 35 (*) "INSMOD-time" means when you load module with
ian@0 36 /sbin/insmod you can feed it optional parameters.
ian@0 37 See "man insmod".
ian@0 38 --------------------------
ian@0 39
ian@0 40
ian@0 41 8390 based Network Modules (Paul Gortmaker, Nov 12, 1995)
ian@0 42 --------------------------
ian@0 43
ian@0 44 (Includes: smc-ultra, ne, wd, 3c503, hp, hp-plus, e2100 and ac3200)
ian@0 45
ian@0 46 The 8390 series of network drivers now support multiple card systems without
ian@0 47 reloading the same module multiple times (memory efficient!) This is done by
ian@0 48 specifying multiple comma separated values, such as:
ian@0 49
ian@0 50 insmod 3c503.o io=0x280,0x300,0x330,0x350 xcvr=0,1,0,1
ian@0 51
ian@0 52 The above would have the one module controlling four 3c503 cards, with card 2
ian@0 53 and 4 using external transceivers. The "insmod" manual describes the usage
ian@0 54 of comma separated value lists.
ian@0 55
ian@0 56 It is *STRONGLY RECOMMENDED* that you supply "io=" instead of autoprobing.
ian@0 57 If an "io=" argument is not supplied, then the ISA drivers will complain
ian@0 58 about autoprobing being not recommended, and begrudgingly autoprobe for
ian@0 59 a *SINGLE CARD ONLY* -- if you want to use multiple cards you *have* to
ian@0 60 supply an "io=0xNNN,0xQQQ,..." argument.
ian@0 61
ian@0 62 The ne module is an exception to the above. A NE2000 is essentially an
ian@0 63 8390 chip, some bus glue and some RAM. Because of this, the ne probe is
ian@0 64 more invasive than the rest, and so at boot we make sure the ne probe is
ian@0 65 done last of all the 8390 cards (so that it won't trip over other 8390 based
ian@0 66 cards) With modules we can't ensure that all other non-ne 8390 cards have
ian@0 67 already been found. Because of this, the ne module REQUIRES an "io=0xNNN"
ian@0 68 argument passed in via insmod. It will refuse to autoprobe.
ian@0 69
ian@0 70 It is also worth noting that auto-IRQ probably isn't as reliable during
ian@0 71 the flurry of interrupt activity on a running machine. Cards such as the
ian@0 72 ne2000 that can't get the IRQ setting from an EEPROM or configuration
ian@0 73 register are probably best supplied with an "irq=M" argument as well.
ian@0 74
ian@0 75
ian@0 76 ----------------------------------------------------------------------
ian@0 77 Card/Module List - Configurable Parameters and Default Values
ian@0 78 ----------------------------------------------------------------------
ian@0 79
ian@0 80 3c501.c:
ian@0 81 io = 0x280 IO base address
ian@0 82 irq = 5 IRQ
ian@0 83 (Probes ports: 0x280, 0x300)
ian@0 84
ian@0 85 3c503.c:
ian@0 86 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 87 irq = 0 (IRQ software selected by driver using autoIRQ)
ian@0 88 xcvr = 0 (Use xcvr=1 to select external transceiver.)
ian@0 89 (Probes ports: 0x300, 0x310, 0x330, 0x350, 0x250, 0x280, 0x2A0, 0x2E0)
ian@0 90
ian@0 91 3c505.c:
ian@0 92 io = 0
ian@0 93 irq = 0
ian@0 94 dma = 6 (not autoprobed)
ian@0 95 (Probes ports: 0x300, 0x280, 0x310)
ian@0 96
ian@0 97 3c507.c:
ian@0 98 io = 0x300
ian@0 99 irq = 0
ian@0 100 (Probes ports: 0x300, 0x320, 0x340, 0x280)
ian@0 101
ian@0 102 3c509.c:
ian@0 103 io = 0
ian@0 104 irq = 0
ian@0 105 ( Module load-time probing Works reliably only on EISA, ISA ID-PROBE
ian@0 106 IS NOT RELIABLE! Compile this driver statically into kernel for
ian@0 107 now, if you need it auto-probing on an ISA-bus machine. )
ian@0 108
ian@0 109 8390.c:
ian@0 110 (No public options, several other modules need this one)
ian@0 111
ian@0 112 a2065.c:
ian@0 113 Since this is a Zorro board, it supports full autoprobing, even for
ian@0 114 multiple boards. (m68k/Amiga)
ian@0 115
ian@0 116 ac3200.c:
ian@0 117 io = 0 (Checks 0x1000 to 0x8fff in 0x1000 intervals)
ian@0 118 irq = 0 (Read from config register)
ian@0 119 (EISA probing..)
ian@0 120
ian@0 121 apricot.c:
ian@0 122 io = 0x300 (Can't be altered!)
ian@0 123 irq = 10
ian@0 124
ian@0 125 arcnet.c:
ian@0 126 io = 0
ian@0 127 irqnum = 0
ian@0 128 shmem = 0
ian@0 129 num = 0
ian@0 130 DO SET THESE MANUALLY AT INSMOD!
ian@0 131 (When probing, looks at the following possible addresses:
ian@0 132 Suggested ones:
ian@0 133 0x300, 0x2E0, 0x2F0, 0x2D0
ian@0 134 Other ones:
ian@0 135 0x200, 0x210, 0x220, 0x230, 0x240, 0x250, 0x260, 0x270,
ian@0 136 0x280, 0x290, 0x2A0, 0x2B0, 0x2C0,
ian@0 137 0x310, 0x320, 0x330, 0x340, 0x350, 0x360, 0x370,
ian@0 138 0x380, 0x390, 0x3A0, 0x3E0, 0x3F0 )
ian@0 139
ian@0 140 ariadne.c:
ian@0 141 Since this is a Zorro board, it supports full autoprobing, even for
ian@0 142 multiple boards. (m68k/Amiga)
ian@0 143
ian@0 144 at1700.c:
ian@0 145 io = 0x260
ian@0 146 irq = 0
ian@0 147 (Probes ports: 0x260, 0x280, 0x2A0, 0x240, 0x340, 0x320, 0x380, 0x300)
ian@0 148
ian@0 149 atari_bionet.c:
ian@0 150 Supports full autoprobing. (m68k/Atari)
ian@0 151
ian@0 152 atari_pamsnet.c:
ian@0 153 Supports full autoprobing. (m68k/Atari)
ian@0 154
ian@0 155 atarilance.c:
ian@0 156 Supports full autoprobing. (m68k/Atari)
ian@0 157
ian@0 158 atp.c: *Not modularized*
ian@0 159 (Probes ports: 0x378, 0x278, 0x3BC;
ian@0 160 fixed IRQs: 5 and 7 )
ian@0 161
ian@0 162 cops.c:
ian@0 163 io = 0x240
ian@0 164 irq = 5
ian@0 165 nodeid = 0 (AutoSelect = 0, NodeID 1-254 is hand selected.)
ian@0 166 (Probes ports: 0x240, 0x340, 0x200, 0x210, 0x220, 0x230, 0x260,
ian@0 167 0x2A0, 0x300, 0x310, 0x320, 0x330, 0x350, 0x360)
ian@0 168
ian@0 169 de4x5.c:
ian@0 170 io = 0x000b
ian@0 171 irq = 10
ian@0 172 is_not_dec = 0 -- For non-DEC card using DEC 21040/21041/21140 chip, set this to 1
ian@0 173 (EISA, and PCI probing)
ian@0 174
ian@0 175 de600.c:
ian@0 176 de600_debug = 0
ian@0 177 (On port 0x378, irq 7 -- lpt1; compile time configurable)
ian@0 178
ian@0 179 de620.c:
ian@0 180 bnc = 0, utp = 0 <-- Force media by setting either.
ian@0 181 io = 0x378 (also compile-time configurable)
ian@0 182 irq = 7
ian@0 183
ian@0 184 depca.c:
ian@0 185 io = 0x200
ian@0 186 irq = 7
ian@0 187 (Probes ports: ISA: 0x300, 0x200;
ian@0 188 EISA: 0x0c00 )
ian@0 189
ian@0 190 dummy.c:
ian@0 191 No options
ian@0 192
ian@0 193 e2100.c:
ian@0 194 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 195 irq = 0 (IRQ software selected by driver)
ian@0 196 mem = 0 (Override default shared memory start of 0xd0000)
ian@0 197 xcvr = 0 (Use xcvr=1 to select external transceiver.)
ian@0 198 (Probes ports: 0x300, 0x280, 0x380, 0x220)
ian@0 199
ian@0 200 eepro.c:
ian@0 201 io = 0x200
ian@0 202 irq = 0
ian@0 203 (Probes ports: 0x200, 0x240, 0x280, 0x2C0, 0x300, 0x320, 0x340, 0x360)
ian@0 204
ian@0 205 eexpress.c:
ian@0 206 io = 0x300
ian@0 207 irq = 0 (IRQ value read from EEPROM)
ian@0 208 (Probes ports: 0x300, 0x270, 0x320, 0x340)
ian@0 209
ian@0 210 eql.c:
ian@0 211 (No parameters)
ian@0 212
ian@0 213 ewrk3.c:
ian@0 214 io = 0x300
ian@0 215 irq = 5
ian@0 216 (With module no autoprobing!
ian@0 217 On EISA-bus does EISA probing.
ian@0 218 Static linkage probes ports on ISA bus:
ian@0 219 0x100, 0x120, 0x140, 0x160, 0x180, 0x1A0, 0x1C0,
ian@0 220 0x200, 0x220, 0x240, 0x260, 0x280, 0x2A0, 0x2C0, 0x2E0,
ian@0 221 0x300, 0x340, 0x360, 0x380, 0x3A0, 0x3C0)
ian@0 222
ian@0 223 hp-plus.c:
ian@0 224 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 225 irq = 0 (IRQ read from configuration register)
ian@0 226 (Probes ports: 0x200, 0x240, 0x280, 0x2C0, 0x300, 0x320, 0x340)
ian@0 227
ian@0 228 hp.c:
ian@0 229 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 230 irq = 0 (IRQ software selected by driver using autoIRQ)
ian@0 231 (Probes ports: 0x300, 0x320, 0x340, 0x280, 0x2C0, 0x200, 0x240)
ian@0 232
ian@0 233 hp100.c:
ian@0 234 hp100_port = 0 (IO-base address)
ian@0 235 (Does EISA-probing, if on EISA-slot;
ian@0 236 On ISA-bus probes all ports from 0x100 thru to 0x3E0
ian@0 237 in increments of 0x020)
ian@0 238
ian@0 239 hydra.c:
ian@0 240 Since this is a Zorro board, it supports full autoprobing, even for
ian@0 241 multiple boards. (m68k/Amiga)
ian@0 242
ian@0 243 ibmtr.c:
ian@0 244 io = 0xa20, 0xa24 (autoprobed by default)
ian@0 245 irq = 0 (driver cannot select irq - read from hardware)
ian@0 246 mem = 0 (shared memory base set at 0xd0000 and not yet
ian@0 247 able to override thru mem= parameter.)
ian@0 248
ian@0 249 lance.c: *Not modularized*
ian@0 250 (PCI, and ISA probing; "CONFIG_PCI" needed for PCI support)
ian@0 251 (Probes ISA ports: 0x300, 0x320, 0x340, 0x360)
ian@0 252
ian@0 253 loopback.c: *Static kernel component*
ian@0 254
ian@0 255 ne.c:
ian@0 256 io = 0 (Explicitly *requires* an "io=0xNNN" value)
ian@0 257 irq = 0 (Tries to determine configured IRQ via autoIRQ)
ian@0 258 (Probes ports: 0x300, 0x280, 0x320, 0x340, 0x360)
ian@0 259
ian@0 260 net_init.c: *Static kernel component*
ian@0 261
ian@0 262 ni52.c: *Not modularized*
ian@0 263 (Probes ports: 0x300, 0x280, 0x360, 0x320, 0x340
ian@0 264 mems: 0xD0000, 0xD2000, 0xC8000, 0xCA000,
ian@0 265 0xD4000, 0xD6000, 0xD8000 )
ian@0 266
ian@0 267 ni65.c: *Not modularized* **16MB MEMORY BARRIER BUG**
ian@0 268 (Probes ports: 0x300, 0x320, 0x340, 0x360)
ian@0 269
ian@0 270 pi2.c: *Not modularized* (well, NON-STANDARD modularization!)
ian@0 271 Only one card supported at this time.
ian@0 272 (Probes ports: 0x380, 0x300, 0x320, 0x340, 0x360, 0x3A0)
ian@0 273
ian@0 274 plip.c:
ian@0 275 io = 0
ian@0 276 irq = 0 (by default, uses IRQ 5 for port at 0x3bc, IRQ 7
ian@0 277 for port at 0x378, and IRQ 2 for port at 0x278)
ian@0 278 (Probes ports: 0x278, 0x378, 0x3bc)
ian@0 279
ian@0 280 ppp.c:
ian@0 281 No options (ppp-2.2+ has some, this is based on non-dynamic
ian@0 282 version from ppp-2.1.2d)
ian@0 283
ian@0 284 seeq8005.c: *Not modularized*
ian@0 285 (Probes ports: 0x300, 0x320, 0x340, 0x360)
ian@0 286
ian@0 287 skeleton.c: *Skeleton*
ian@0 288
ian@0 289 slhc.c:
ian@0 290 No configuration parameters
ian@0 291
ian@0 292 slip.c:
ian@0 293 slip_maxdev = 256 (default value from SL_NRUNIT on slip.h)
ian@0 294
ian@0 295
ian@0 296 smc-ultra.c:
ian@0 297 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 298 irq = 0 (IRQ val. read from EEPROM)
ian@0 299 (Probes ports: 0x200, 0x220, 0x240, 0x280, 0x300, 0x340, 0x380)
ian@0 300
ian@0 301 tulip.c: *Partial modularization*
ian@0 302 (init-time memory allocation makes problems..)
ian@0 303
ian@0 304 tunnel.c:
ian@0 305 No insmod parameters
ian@0 306
ian@0 307 wavelan.c:
ian@0 308 io = 0x390 (Settable, but change not recommended)
ian@0 309 irq = 0 (Not honoured, if changed..)
ian@0 310
ian@0 311 wd.c:
ian@0 312 io = 0 (It will complain if you don't supply an "io=0xNNN")
ian@0 313 irq = 0 (IRQ val. read from EEPROM, ancient cards use autoIRQ)
ian@0 314 mem = 0 (Force shared-memory on address 0xC8000, or whatever..)
ian@0 315 mem_end = 0 (Force non-std. mem. size via supplying mem_end val.)
ian@0 316 (eg. for 32k WD8003EBT, use mem=0xd0000 mem_end=0xd8000)
ian@0 317 (Probes ports: 0x300, 0x280, 0x380, 0x240)
ian@0 318
ian@0 319 znet.c: *Not modularized*
ian@0 320 (Only one device on Zenith Z-Note (notebook?) systems,
ian@0 321 configuration information from (EE)PROM)