2 # For a description of the syntax of this configuration file,
3 # see Documentation/kbuild/kconfig-language.txt.
10 config LOCKDEP_SUPPORT
14 config STACKTRACE_SUPPORT
18 config RWSEM_GENERIC_SPINLOCK
21 config RWSEM_XCHGADD_ALGORITHM
25 config ARCH_HAS_ILOG2_U32
29 config ARCH_HAS_ILOG2_U64
33 config GENERIC_HWEIGHT
40 mainmenu "Linux Kernel Configuration"
50 comment "Processor type and features"
55 Select this option if you have a 64 bit IBM zSeries machine
56 and want to use the 64 bit addressing mode.
63 bool "Symmetric multi-processing support"
65 This enables support for systems with more than one CPU. If you have
66 a system with only one CPU, like most personal computers, say N. If
67 you have a system with more than one CPU, say Y.
69 If you say N here, the kernel will run on single and multiprocessor
70 machines, but will use only one CPU of a multiprocessor machine. If
71 you say Y here, the kernel will run on many, but not all,
72 singleprocessor machines. On a singleprocessor machine, the kernel
73 will run faster if you say N here.
75 See also the <file:Documentation/smp.txt> and the SMP-HOWTO
76 available at <http://www.tldp.org/docs.html#howto>.
78 Even if you don't know what to do here, say Y.
81 int "Maximum number of CPUs (2-64)"
86 This allows you to specify the maximum number of CPUs which this
87 kernel will support. The maximum supported value is 64 and the
88 minimum value which makes sense is 2.
90 This is purely to save memory - each supported CPU adds
91 approximately sixteen kilobytes to the kernel image.
94 bool "Support for hot-pluggable CPUs"
99 Say Y here to be able to turn CPUs off and on. CPUs
100 can be controlled through /sys/devices/system/cpu/cpu#.
101 Say N if you want to disable CPU hotplug.
103 config DEFAULT_MIGRATION_COST
108 bool "IEEE FPU emulation"
111 This option is required for IEEE compliant floating point arithmetic
112 on older S/390 machines. Say Y unless you know your machine doesn't
116 bool "Kernel support for 31 bit emulation"
119 Select this option if you want to enable your system kernel to
120 handle system-calls from ELF binaries for 31 bit ESA. This option
121 (and some other stuff like libraries and such) is needed for
122 executing 31 bit applications. It is safe to say "Y".
124 config SYSVIPC_COMPAT
126 depends on COMPAT && SYSVIPC
133 config S390_SWITCH_AMODE
134 bool "Switch kernel/user addressing modes"
136 This option allows to switch the addressing modes of kernel and user
137 space. The kernel parameter switch_amode=on will enable this feature,
138 default is disabled. Enabling this (via kernel parameter) on machines
139 earlier than IBM System z9-109 EC/BC will reduce system performance.
141 Note that this option will also be selected by selecting the execute
142 protection option below. Enabling the execute protection via the
143 noexec kernel parameter will also switch the addressing modes,
144 independent of the switch_amode kernel parameter.
147 config S390_EXEC_PROTECT
148 bool "Data execute protection"
149 select S390_SWITCH_AMODE
151 This option allows to enable a buffer overflow protection for user
152 space programs and it also selects the addressing mode option above.
153 The kernel parameter noexec=on will enable this feature and also
154 switch the addressing modes, default is disabled. Enabling this (via
155 kernel parameter) on machines earlier than IBM System z9-109 EC/BC
156 will reduce system performance.
158 comment "Code generation options"
161 prompt "Processor type"
165 bool "S/390 model G5 and G6"
168 Select this to build a 31 bit kernel that works
169 on all S/390 and zSeries machines.
172 bool "IBM eServer zSeries model z800 and z900"
174 Select this to optimize for zSeries machines. This
175 will enable some optimizations that are not available
176 on older 31 bit only CPUs.
179 bool "IBM eServer zSeries model z890 and z990"
181 Select this enable optimizations for model z890/z990.
182 This will be slightly faster but does not work on
183 older machines such as the z900.
188 Select this to enable optimizations for IBM System z9-109, IBM
189 System z9 Enterprise Class (z9 EC), and IBM System z9 Business
190 Class (z9 BC). The kernel will be slightly faster but will not
191 work on older machines such as the z990, z890, z900, and z800.
196 bool "Pack kernel stack"
198 This option enables the compiler option -mkernel-backchain if it
199 is available. If the option is available the compiler supports
200 the new stack layout which dramatically reduces the minimum stack
201 frame size. With an old compiler a non-leaf function needs a
202 minimum of 96 bytes on 31 bit and 160 bytes on 64 bit. With
203 -mkernel-backchain the minimum size drops to 16 byte on 31 bit
204 and 24 byte on 64 bit.
206 Say Y if you are unsure.
209 bool "Use 4kb/8kb for kernel stack instead of 8kb/16kb"
210 depends on PACK_STACK && !LOCKDEP
212 If you say Y here and the compiler supports the -mkernel-backchain
213 option the kernel will use a smaller kernel stack size. For 31 bit
214 the reduced size is 4kb instead of 8kb and for 64 bit it is 8kb
215 instead of 16kb. This allows to run more thread on a system and
216 reduces the pressure on the memory management for higher order
219 Say N if you are unsure.
223 bool "Detect kernel stack overflow"
225 This option enables the compiler option -mstack-guard and
226 -mstack-size if they are available. If the compiler supports them
227 it will emit additional code to each function prolog to trigger
228 an illegal operation if the kernel stack is about to overflow.
230 Say N if you are unsure.
233 int "Size of the guard area (128-1024)"
235 depends on CHECK_STACK
238 This allows you to specify the size of the guard area at the lower
239 end of the kernel stack. If the kernel stack points into the guard
240 area on function entry an illegal operation is triggered. The size
241 needs to be a power of 2. Please keep in mind that the size of an
242 interrupt frame is 184 bytes for 31 bit and 328 bytes on 64 bit.
243 The minimum size for the stack guard should be 256 for 31 bit and
247 bool "Emit compiler warnings for function with broken stack usage"
249 This option enables the compiler options -mwarn-framesize and
250 -mwarn-dynamicstack. If the compiler supports these options it
251 will generate warnings for function which either use alloca or
252 create a stack frame bigger then CONFIG_WARN_STACK_SIZE.
254 Say N if you are unsure.
256 config WARN_STACK_SIZE
257 int "Maximum frame size considered safe (128-2048)"
259 depends on WARN_STACK
262 This allows you to specify the maximum frame size a function may
263 have without the compiler complaining about it.
265 config ARCH_POPULATES_NODE_MAP
273 comment "I/O subsystem configuration"
275 config MACHCHK_WARNING
276 bool "Process warning machine checks"
278 Select this option if you want the machine check handler on IBM S/390 or
279 zSeries to process warning machine checks (e.g. on power failures).
283 tristate "QDIO support"
285 This driver provides the Queued Direct I/O base support for
288 For details please refer to the documentation provided by IBM at
289 <http://www10.software.ibm.com/developerworks/opensource/linux390>
291 To compile this driver as a module, choose M here: the
292 module will be called qdio.
297 bool "Extended debugging information"
300 Say Y here to get extended debugging output in
301 /sys/kernel/debug/s390dbf/qdio...
302 Warning: this option reduces the performance of the QDIO module.
309 bool "Preemptible Kernel"
311 This option reduces the latency of the kernel when reacting to
312 real-time or interactive events by allowing a low priority process to
313 be preempted even if it is in kernel mode executing a system call.
314 This allows applications to run more reliably even when the system is
317 Say N if you are unsure.
320 bool "Builtin IPL record support"
322 If you want to use the produced kernel to IPL directly from a
323 device, you have to merge a bootsector specific to the device
324 into the first bytes of the kernel. You will have to select the
328 prompt "IPL method generated into head.S"
332 Select "tape" if you want to IPL the image from a Tape.
334 Select "vm_reader" if you are running under VM/ESA and want
335 to IPL the image from the emulated card reader.
345 source "fs/Kconfig.binfmt"
348 bool "Show crashed user process info"
350 Say Y to print all process fault locations to the console. This is
351 a debugging option; you probably do not want to set it unless you
352 are an S390 port maintainer.
355 bool "Pseudo page fault support"
357 Select this option, if you want to use PFAULT pseudo page fault
358 handling under VM. If running native or in LPAR, this option
359 has no effect. If your VM does not support PFAULT, PAGEEX
360 pseudo page fault handling will be used.
361 Note that VM 4.2 supports PFAULT but has a bug in its
362 implementation that causes some problems.
363 Everybody who wants to run Linux under VM != VM4.2 should select
367 bool "VM shared kernel support"
369 Select this option, if you want to share the text segment of the
370 Linux kernel between different VM guests. This reduces memory
371 usage with lots of guests but greatly increases kernel size.
372 You should only select this option if you know what you are
373 doing and want to exploit this feature.
376 tristate "Cooperative memory management"
378 Select this option, if you want to enable the kernel interface
379 to reduce the memory size of the system. This is accomplished
380 by allocating pages of memory and put them "on hold". This only
381 makes sense for a system running under VM where the unused pages
382 will be reused by VM for other guest systems. The interface
383 allows an external monitor to balance memory of many systems.
384 Everybody who wants to run Linux under VM should select this
388 bool "/proc interface to cooperative memory management"
391 Select this option to enable the /proc interface to the
392 cooperative memory management.
395 bool "IUCV special message interface to cooperative memory management"
396 depends on CMM && (SMSGIUCV=y || CMM=SMSGIUCV)
398 Select this option to enable the special message interface to
399 the cooperative memory management.
402 bool "Virtual CPU timer support"
404 This provides a kernel interface for virtual CPU timers.
407 config VIRT_CPU_ACCOUNTING
408 bool "Base user process accounting on virtual cpu timer"
409 depends on VIRT_TIMER
411 Select this option to use CPU timer deltas to do user
415 bool "Linux - VM Monitor Stream, base infrastructure"
416 depends on PROC_FS && VIRT_TIMER=y
418 This provides a kernel interface for creating and updating z/VM APPLDATA
419 monitor records. The monitor records are updated at certain time
420 intervals, once the timer is started.
421 Writing 1 or 0 to /proc/appldata/timer starts(1) or stops(0) the timer,
422 i.e. enables or disables monitoring on the Linux side.
423 A custom interval value (in seconds) can be written to
424 /proc/appldata/interval.
426 Defaults are 60 seconds interval and timer off.
427 The /proc entries can also be read from, showing the current settings.
430 tristate "Monitor memory management statistics"
431 depends on APPLDATA_BASE && VM_EVENT_COUNTERS
433 This provides memory management related data to the Linux - VM Monitor
434 Stream, like paging/swapping rate, memory utilisation, etc.
435 Writing 1 or 0 to /proc/appldata/memory creates(1) or removes(0) a z/VM
436 APPLDATA monitor record, i.e. enables or disables monitoring this record
440 The /proc entry can also be read from, showing the current settings.
442 This can also be compiled as a module, which will be called
446 tristate "Monitor OS statistics"
447 depends on APPLDATA_BASE
449 This provides OS related data to the Linux - VM Monitor Stream, like
450 CPU utilisation, etc.
451 Writing 1 or 0 to /proc/appldata/os creates(1) or removes(0) a z/VM
452 APPLDATA monitor record, i.e. enables or disables monitoring this record
456 This can also be compiled as a module, which will be called
459 config APPLDATA_NET_SUM
460 tristate "Monitor overall network statistics"
461 depends on APPLDATA_BASE
463 This provides network related data to the Linux - VM Monitor Stream,
464 currently there is only a total sum of network I/O statistics, no
466 Writing 1 or 0 to /proc/appldata/net_sum creates(1) or removes(0) a z/VM
467 APPLDATA monitor record, i.e. enables or disables monitoring this record
471 This can also be compiled as a module, which will be called
475 bool "No HZ timer ticks in idle"
477 Switches the regular HZ timer off when the system is going idle.
478 This helps z/VM to detect that the Linux system is idle. VM can
479 then "swap-out" this guest which reduces memory usage. It also
480 reduces the overhead of idle systems.
482 The HZ timer can be switched on/off via /proc/sys/kernel/hz_timer.
483 hz_timer=0 means HZ timer is disabled. hz_timer=1 means HZ
486 config NO_IDLE_HZ_INIT
487 bool "HZ timer in idle off by default"
488 depends on NO_IDLE_HZ
490 The HZ timer is switched off in idle by default. That means the
491 HZ timer is already disabled at boot time.
494 bool "s390 hypervisor file system support"
495 select SYS_HYPERVISOR
498 This is a virtual file system intended to provide accounting
499 information in an s390 hypervisor environment.
502 bool "kexec system call"
504 kexec is a system call that implements the ability to shutdown your
505 current kernel, and to start another kernel. It is like a reboot
506 but is independent of hardware/microcode support.
516 source "drivers/base/Kconfig"
518 source "drivers/connector/Kconfig"
520 source "drivers/scsi/Kconfig"
522 source "drivers/s390/Kconfig"
524 source "drivers/net/Kconfig"
528 menu "Instrumentation Support"
530 source "arch/s390/oprofile/Kconfig"
533 bool "Kprobes (EXPERIMENTAL)"
534 depends on EXPERIMENTAL && MODULES
536 Kprobes allows you to trap at almost any kernel address and
537 execute a callback function. register_kprobe() establishes
538 a probepoint and specifies the callback. Kprobes is useful
539 for kernel debugging, non-intrusive instrumentation and testing.
540 If in doubt, say "N".
544 source "arch/s390/Kconfig.debug"
546 source "security/Kconfig"
548 source "crypto/Kconfig"