1config ARCH 2 string 3 option env="ARCH" 4 5config KERNELVERSION 6 string 7 option env="KERNELVERSION" 8 9config DEFCONFIG_LIST 10 string 11 depends on !UML 12 option defconfig_list 13 default "/lib/modules/$UNAME_RELEASE/.config" 14 default "/etc/kernel-config" 15 default "/boot/config-$UNAME_RELEASE" 16 default "$ARCH_DEFCONFIG" 17 default "arch/$ARCH/defconfig" 18 19config CONSTRUCTORS 20 bool 21 depends on !UML 22 23config IRQ_WORK 24 bool 25 26config BUILDTIME_EXTABLE_SORT 27 bool 28 29menu "General setup" 30 31config BROKEN 32 bool 33 34config BROKEN_ON_SMP 35 bool 36 depends on BROKEN || !SMP 37 default y 38 39config INIT_ENV_ARG_LIMIT 40 int 41 default 32 if !UML 42 default 128 if UML 43 help 44 Maximum of each of the number of arguments and environment 45 variables passed to init from the kernel command line. 46 47 48config CROSS_COMPILE 49 string "Cross-compiler tool prefix" 50 help 51 Same as running 'make CROSS_COMPILE=prefix-' but stored for 52 default make runs in this kernel build directory. You don't 53 need to set this unless you want the configured kernel build 54 directory to select the cross-compiler automatically. 55 56config COMPILE_TEST 57 bool "Compile also drivers which will not load" 58 default n 59 help 60 Some drivers can be compiled on a different platform than they are 61 intended to be run on. Despite they cannot be loaded there (or even 62 when they load they cannot be used due to missing HW support), 63 developers still, opposing to distributors, might want to build such 64 drivers to compile-test them. 65 66 If you are a developer and want to build everything available, say Y 67 here. If you are a user/distributor, say N here to exclude useless 68 drivers to be distributed. 69 70config LOCALVERSION 71 string "Local version - append to kernel release" 72 help 73 Append an extra string to the end of your kernel version. 74 This will show up when you type uname, for example. 75 The string you set here will be appended after the contents of 76 any files with a filename matching localversion* in your 77 object and source tree, in that order. Your total string can 78 be a maximum of 64 characters. 79 80config LOCALVERSION_AUTO 81 bool "Automatically append version information to the version string" 82 default y 83 help 84 This will try to automatically determine if the current tree is a 85 release tree by looking for git tags that belong to the current 86 top of tree revision. 87 88 A string of the format -gxxxxxxxx will be added to the localversion 89 if a git-based tree is found. The string generated by this will be 90 appended after any matching localversion* files, and after the value 91 set in CONFIG_LOCALVERSION. 92 93 (The actual string used here is the first eight characters produced 94 by running the command: 95 96 $ git rev-parse --verify HEAD 97 98 which is done within the script "scripts/setlocalversion".) 99 100config HAVE_KERNEL_GZIP 101 bool 102 103config HAVE_KERNEL_BZIP2 104 bool 105 106config HAVE_KERNEL_LZMA 107 bool 108 109config HAVE_KERNEL_XZ 110 bool 111 112config HAVE_KERNEL_LZO 113 bool 114 115config HAVE_KERNEL_LZ4 116 bool 117 118choice 119 prompt "Kernel compression mode" 120 default KERNEL_GZIP 121 help 122 The linux kernel is a kind of self-extracting executable. 123 Several compression algorithms are available, which differ 124 in efficiency, compression and decompression speed. 125 Compression speed is only relevant when building a kernel. 126 Decompression speed is relevant at each boot. 127 128 If you have any problems with bzip2 or lzma compressed 129 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 130 version of this functionality (bzip2 only), for 2.4, was 131 supplied by Christian Ludwig) 132 133 High compression options are mostly useful for users, who 134 are low on disk space (embedded systems), but for whom ram 135 size matters less. 136 137 If in doubt, select 'gzip' 138 139config KERNEL_UNCOMPRESSED 140 bool "No compression" 141 help 142 No compression at all. The kernel is huge but the compression and 143 decompression times are zero. 144 This is usually not what you want. 145 146config KERNEL_GZIP 147 bool "Gzip" 148 depends on HAVE_KERNEL_GZIP 149 help 150 The old and tried gzip compression. It provides a good balance 151 between compression ratio and decompression speed. 152 153config KERNEL_BZIP2 154 bool "Bzip2" 155 depends on HAVE_KERNEL_BZIP2 156 help 157 Its compression ratio and speed is intermediate. 158 Decompression speed is slowest among the choices. The kernel 159 size is about 10% smaller with bzip2, in comparison to gzip. 160 Bzip2 uses a large amount of memory. For modern kernels you 161 will need at least 8MB RAM or more for booting. 162 163config KERNEL_LZMA 164 bool "LZMA" 165 depends on HAVE_KERNEL_LZMA 166 help 167 This compression algorithm's ratio is best. Decompression speed 168 is between gzip and bzip2. Compression is slowest. 169 The kernel size is about 33% smaller with LZMA in comparison to gzip. 170 171config KERNEL_XZ 172 bool "XZ" 173 depends on HAVE_KERNEL_XZ 174 help 175 XZ uses the LZMA2 algorithm and instruction set specific 176 BCJ filters which can improve compression ratio of executable 177 code. The size of the kernel is about 30% smaller with XZ in 178 comparison to gzip. On architectures for which there is a BCJ 179 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 180 will create a few percent smaller kernel than plain LZMA. 181 182 The speed is about the same as with LZMA: The decompression 183 speed of XZ is better than that of bzip2 but worse than gzip 184 and LZO. Compression is slow. 185 186config KERNEL_LZO 187 bool "LZO" 188 depends on HAVE_KERNEL_LZO 189 help 190 Its compression ratio is the poorest among the choices. The kernel 191 size is about 10% bigger than gzip; however its speed 192 (both compression and decompression) is the fastest. 193 194config KERNEL_LZ4 195 bool "LZ4" 196 depends on HAVE_KERNEL_LZ4 197 help 198 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 199 A preliminary version of LZ4 de/compression tool is available at 200 <https://code.google.com/p/lz4/>. 201 202 Its compression ratio is worse than LZO. The size of the kernel 203 is about 8% bigger than LZO. But the decompression speed is 204 faster than LZO. 205 206endchoice 207 208config DEFAULT_HOSTNAME 209 string "Default hostname" 210 default "(none)" 211 help 212 This option determines the default system hostname before userspace 213 calls sethostname(2). The kernel traditionally uses "(none)" here, 214 but you may wish to use a different default here to make a minimal 215 system more usable with less configuration. 216 217config SWAP 218 bool "Support for paging of anonymous memory (swap)" 219 depends on MMU && BLOCK 220 default y 221 help 222 This option allows you to choose whether you want to have support 223 for so called swap devices or swap files in your kernel that are 224 used to provide more virtual memory than the actual RAM present 225 in your computer. If unsure say Y. 226 227config SYSVIPC 228 bool "System V IPC" 229 ---help--- 230 Inter Process Communication is a suite of library functions and 231 system calls which let processes (running programs) synchronize and 232 exchange information. It is generally considered to be a good thing, 233 and some programs won't run unless you say Y here. In particular, if 234 you want to run the DOS emulator dosemu under Linux (read the 235 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 236 you'll need to say Y here. 237 238 You can find documentation about IPC with "info ipc" and also in 239 section 6.4 of the Linux Programmer's Guide, available from 240 <http://www.tldp.org/guides.html>. 241 242config SYSVIPC_SYSCTL 243 bool 244 depends on SYSVIPC 245 depends on SYSCTL 246 default y 247 248config POSIX_MQUEUE 249 bool "POSIX Message Queues" 250 depends on NET 251 ---help--- 252 POSIX variant of message queues is a part of IPC. In POSIX message 253 queues every message has a priority which decides about succession 254 of receiving it by a process. If you want to compile and run 255 programs written e.g. for Solaris with use of its POSIX message 256 queues (functions mq_*) say Y here. 257 258 POSIX message queues are visible as a filesystem called 'mqueue' 259 and can be mounted somewhere if you want to do filesystem 260 operations on message queues. 261 262 If unsure, say Y. 263 264config POSIX_MQUEUE_SYSCTL 265 bool 266 depends on POSIX_MQUEUE 267 depends on SYSCTL 268 default y 269 270config FHANDLE 271 bool "open by fhandle syscalls" 272 select EXPORTFS 273 help 274 If you say Y here, a user level program will be able to map 275 file names to handle and then later use the handle for 276 different file system operations. This is useful in implementing 277 userspace file servers, which now track files using handles instead 278 of names. The handle would remain the same even if file names 279 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 280 syscalls. 281 282config AUDIT 283 bool "Auditing support" 284 depends on NET 285 help 286 Enable auditing infrastructure that can be used with another 287 kernel subsystem, such as SELinux (which requires this for 288 logging of avc messages output). Does not do system-call 289 auditing without CONFIG_AUDITSYSCALL. 290 291config AUDITSYSCALL 292 bool "Enable system-call auditing support" 293 depends on AUDIT && (X86 || PARISC || PPC || S390 || IA64 || UML || SPARC64 || SUPERH || (ARM && AEABI && !OABI_COMPAT)) 294 default y if SECURITY_SELINUX 295 help 296 Enable low-overhead system-call auditing infrastructure that 297 can be used independently or with another kernel subsystem, 298 such as SELinux. 299 300config AUDIT_WATCH 301 def_bool y 302 depends on AUDITSYSCALL 303 select FSNOTIFY 304 305config AUDIT_TREE 306 def_bool y 307 depends on AUDITSYSCALL 308 select FSNOTIFY 309 310config AUDIT_LOGINUID_IMMUTABLE 311 bool "Make audit loginuid immutable" 312 depends on AUDIT 313 help 314 The config option toggles if a task setting its loginuid requires 315 CAP_SYS_AUDITCONTROL or if that task should require no special permissions 316 but should instead only allow setting its loginuid if it was never 317 previously set. On systems which use systemd or a similar central 318 process to restart login services this should be set to true. On older 319 systems in which an admin would typically have to directly stop and 320 start processes this should be set to false. Setting this to true allows 321 one to drop potentially dangerous capabilites from the login tasks, 322 but may not be backwards compatible with older init systems. 323 324source "kernel/irq/Kconfig" 325source "kernel/time/Kconfig" 326 327menu "CPU/Task time and stats accounting" 328 329config VIRT_CPU_ACCOUNTING 330 bool 331 332choice 333 prompt "Cputime accounting" 334 default TICK_CPU_ACCOUNTING if !PPC64 335 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 336 337# Kind of a stub config for the pure tick based cputime accounting 338config TICK_CPU_ACCOUNTING 339 bool "Simple tick based cputime accounting" 340 depends on !S390 && !NO_HZ_FULL 341 help 342 This is the basic tick based cputime accounting that maintains 343 statistics about user, system and idle time spent on per jiffies 344 granularity. 345 346 If unsure, say Y. 347 348config VIRT_CPU_ACCOUNTING_NATIVE 349 bool "Deterministic task and CPU time accounting" 350 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 351 select VIRT_CPU_ACCOUNTING 352 help 353 Select this option to enable more accurate task and CPU time 354 accounting. This is done by reading a CPU counter on each 355 kernel entry and exit and on transitions within the kernel 356 between system, softirq and hardirq state, so there is a 357 small performance impact. In the case of s390 or IBM POWER > 5, 358 this also enables accounting of stolen time on logically-partitioned 359 systems. 360 361config VIRT_CPU_ACCOUNTING_GEN 362 bool "Full dynticks CPU time accounting" 363 depends on HAVE_CONTEXT_TRACKING 364 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 365 select VIRT_CPU_ACCOUNTING 366 select CONTEXT_TRACKING 367 help 368 Select this option to enable task and CPU time accounting on full 369 dynticks systems. This accounting is implemented by watching every 370 kernel-user boundaries using the context tracking subsystem. 371 The accounting is thus performed at the expense of some significant 372 overhead. 373 374 For now this is only useful if you are working on the full 375 dynticks subsystem development. 376 377 If unsure, say N. 378 379config IRQ_TIME_ACCOUNTING 380 bool "Fine granularity task level IRQ time accounting" 381 depends on HAVE_IRQ_TIME_ACCOUNTING && !NO_HZ_FULL 382 help 383 Select this option to enable fine granularity task irq time 384 accounting. This is done by reading a timestamp on each 385 transitions between softirq and hardirq state, so there can be a 386 small performance impact. 387 388 If in doubt, say N here. 389 390endchoice 391 392config BSD_PROCESS_ACCT 393 bool "BSD Process Accounting" 394 help 395 If you say Y here, a user level program will be able to instruct the 396 kernel (via a special system call) to write process accounting 397 information to a file: whenever a process exits, information about 398 that process will be appended to the file by the kernel. The 399 information includes things such as creation time, owning user, 400 command name, memory usage, controlling terminal etc. (the complete 401 list is in the struct acct in <file:include/linux/acct.h>). It is 402 up to the user level program to do useful things with this 403 information. This is generally a good idea, so say Y. 404 405config BSD_PROCESS_ACCT_V3 406 bool "BSD Process Accounting version 3 file format" 407 depends on BSD_PROCESS_ACCT 408 default n 409 help 410 If you say Y here, the process accounting information is written 411 in a new file format that also logs the process IDs of each 412 process and it's parent. Note that this file format is incompatible 413 with previous v0/v1/v2 file formats, so you will need updated tools 414 for processing it. A preliminary version of these tools is available 415 at <http://www.gnu.org/software/acct/>. 416 417config TASKSTATS 418 bool "Export task/process statistics through netlink" 419 depends on NET 420 default n 421 help 422 Export selected statistics for tasks/processes through the 423 generic netlink interface. Unlike BSD process accounting, the 424 statistics are available during the lifetime of tasks/processes as 425 responses to commands. Like BSD accounting, they are sent to user 426 space on task exit. 427 428 Say N if unsure. 429 430config TASK_DELAY_ACCT 431 bool "Enable per-task delay accounting" 432 depends on TASKSTATS 433 help 434 Collect information on time spent by a task waiting for system 435 resources like cpu, synchronous block I/O completion and swapping 436 in pages. Such statistics can help in setting a task's priorities 437 relative to other tasks for cpu, io, rss limits etc. 438 439 Say N if unsure. 440 441config TASK_XACCT 442 bool "Enable extended accounting over taskstats" 443 depends on TASKSTATS 444 help 445 Collect extended task accounting data and send the data 446 to userland for processing over the taskstats interface. 447 448 Say N if unsure. 449 450config TASK_IO_ACCOUNTING 451 bool "Enable per-task storage I/O accounting" 452 depends on TASK_XACCT 453 help 454 Collect information on the number of bytes of storage I/O which this 455 task has caused. 456 457 Say N if unsure. 458 459endmenu # "CPU/Task time and stats accounting" 460 461menu "RCU Subsystem" 462 463choice 464 prompt "RCU Implementation" 465 default TREE_RCU 466 467config TREE_RCU 468 bool "Tree-based hierarchical RCU" 469 depends on !PREEMPT && SMP 470 select IRQ_WORK 471 help 472 This option selects the RCU implementation that is 473 designed for very large SMP system with hundreds or 474 thousands of CPUs. It also scales down nicely to 475 smaller systems. 476 477config TREE_PREEMPT_RCU 478 bool "Preemptible tree-based hierarchical RCU" 479 depends on PREEMPT 480 select IRQ_WORK 481 help 482 This option selects the RCU implementation that is 483 designed for very large SMP systems with hundreds or 484 thousands of CPUs, but for which real-time response 485 is also required. It also scales down nicely to 486 smaller systems. 487 488 Select this option if you are unsure. 489 490config TINY_RCU 491 bool "UP-only small-memory-footprint RCU" 492 depends on !PREEMPT && !SMP 493 help 494 This option selects the RCU implementation that is 495 designed for UP systems from which real-time response 496 is not required. This option greatly reduces the 497 memory footprint of RCU. 498 499endchoice 500 501config PREEMPT_RCU 502 def_bool TREE_PREEMPT_RCU 503 help 504 This option enables preemptible-RCU code that is common between 505 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations. 506 507config RCU_STALL_COMMON 508 def_bool ( TREE_RCU || TREE_PREEMPT_RCU || RCU_TRACE ) 509 help 510 This option enables RCU CPU stall code that is common between 511 the TINY and TREE variants of RCU. The purpose is to allow 512 the tiny variants to disable RCU CPU stall warnings, while 513 making these warnings mandatory for the tree variants. 514 515config CONTEXT_TRACKING 516 bool 517 518config RCU_USER_QS 519 bool "Consider userspace as in RCU extended quiescent state" 520 depends on HAVE_CONTEXT_TRACKING && SMP 521 select CONTEXT_TRACKING 522 help 523 This option sets hooks on kernel / userspace boundaries and 524 puts RCU in extended quiescent state when the CPU runs in 525 userspace. It means that when a CPU runs in userspace, it is 526 excluded from the global RCU state machine and thus doesn't 527 try to keep the timer tick on for RCU. 528 529 Unless you want to hack and help the development of the full 530 dynticks mode, you shouldn't enable this option. It also 531 adds unnecessary overhead. 532 533 If unsure say N 534 535config CONTEXT_TRACKING_FORCE 536 bool "Force context tracking" 537 depends on CONTEXT_TRACKING 538 default y if !NO_HZ_FULL 539 help 540 The major pre-requirement for full dynticks to work is to 541 support the context tracking subsystem. But there are also 542 other dependencies to provide in order to make the full 543 dynticks working. 544 545 This option stands for testing when an arch implements the 546 context tracking backend but doesn't yet fullfill all the 547 requirements to make the full dynticks feature working. 548 Without the full dynticks, there is no way to test the support 549 for context tracking and the subsystems that rely on it: RCU 550 userspace extended quiescent state and tickless cputime 551 accounting. This option copes with the absence of the full 552 dynticks subsystem by forcing the context tracking on all 553 CPUs in the system. 554 555 Say Y only if you're working on the developpement of an 556 architecture backend for the context tracking. 557 558 Say N otherwise, this option brings an overhead that you 559 don't want in production. 560 561 562config RCU_FANOUT 563 int "Tree-based hierarchical RCU fanout value" 564 range 2 64 if 64BIT 565 range 2 32 if !64BIT 566 depends on TREE_RCU || TREE_PREEMPT_RCU 567 default 64 if 64BIT 568 default 32 if !64BIT 569 help 570 This option controls the fanout of hierarchical implementations 571 of RCU, allowing RCU to work efficiently on machines with 572 large numbers of CPUs. This value must be at least the fourth 573 root of NR_CPUS, which allows NR_CPUS to be insanely large. 574 The default value of RCU_FANOUT should be used for production 575 systems, but if you are stress-testing the RCU implementation 576 itself, small RCU_FANOUT values allow you to test large-system 577 code paths on small(er) systems. 578 579 Select a specific number if testing RCU itself. 580 Take the default if unsure. 581 582config RCU_FANOUT_LEAF 583 int "Tree-based hierarchical RCU leaf-level fanout value" 584 range 2 RCU_FANOUT if 64BIT 585 range 2 RCU_FANOUT if !64BIT 586 depends on TREE_RCU || TREE_PREEMPT_RCU 587 default 16 588 help 589 This option controls the leaf-level fanout of hierarchical 590 implementations of RCU, and allows trading off cache misses 591 against lock contention. Systems that synchronize their 592 scheduling-clock interrupts for energy-efficiency reasons will 593 want the default because the smaller leaf-level fanout keeps 594 lock contention levels acceptably low. Very large systems 595 (hundreds or thousands of CPUs) will instead want to set this 596 value to the maximum value possible in order to reduce the 597 number of cache misses incurred during RCU's grace-period 598 initialization. These systems tend to run CPU-bound, and thus 599 are not helped by synchronized interrupts, and thus tend to 600 skew them, which reduces lock contention enough that large 601 leaf-level fanouts work well. 602 603 Select a specific number if testing RCU itself. 604 605 Select the maximum permissible value for large systems. 606 607 Take the default if unsure. 608 609config RCU_FANOUT_EXACT 610 bool "Disable tree-based hierarchical RCU auto-balancing" 611 depends on TREE_RCU || TREE_PREEMPT_RCU 612 default n 613 help 614 This option forces use of the exact RCU_FANOUT value specified, 615 regardless of imbalances in the hierarchy. This is useful for 616 testing RCU itself, and might one day be useful on systems with 617 strong NUMA behavior. 618 619 Without RCU_FANOUT_EXACT, the code will balance the hierarchy. 620 621 Say N if unsure. 622 623config RCU_FAST_NO_HZ 624 bool "Accelerate last non-dyntick-idle CPU's grace periods" 625 depends on NO_HZ_COMMON && SMP 626 default n 627 help 628 This option permits CPUs to enter dynticks-idle state even if 629 they have RCU callbacks queued, and prevents RCU from waking 630 these CPUs up more than roughly once every four jiffies (by 631 default, you can adjust this using the rcutree.rcu_idle_gp_delay 632 parameter), thus improving energy efficiency. On the other 633 hand, this option increases the duration of RCU grace periods, 634 for example, slowing down synchronize_rcu(). 635 636 Say Y if energy efficiency is critically important, and you 637 don't care about increased grace-period durations. 638 639 Say N if you are unsure. 640 641config TREE_RCU_TRACE 642 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU ) 643 select DEBUG_FS 644 help 645 This option provides tracing for the TREE_RCU and 646 TREE_PREEMPT_RCU implementations, permitting Makefile to 647 trivially select kernel/rcutree_trace.c. 648 649config RCU_BOOST 650 bool "Enable RCU priority boosting" 651 depends on RT_MUTEXES && PREEMPT_RCU 652 default n 653 help 654 This option boosts the priority of preempted RCU readers that 655 block the current preemptible RCU grace period for too long. 656 This option also prevents heavy loads from blocking RCU 657 callback invocation for all flavors of RCU. 658 659 Say Y here if you are working with real-time apps or heavy loads 660 Say N here if you are unsure. 661 662config RCU_BOOST_PRIO 663 int "Real-time priority to boost RCU readers to" 664 range 1 99 665 depends on RCU_BOOST 666 default 1 667 help 668 This option specifies the real-time priority to which long-term 669 preempted RCU readers are to be boosted. If you are working 670 with a real-time application that has one or more CPU-bound 671 threads running at a real-time priority level, you should set 672 RCU_BOOST_PRIO to a priority higher then the highest-priority 673 real-time CPU-bound thread. The default RCU_BOOST_PRIO value 674 of 1 is appropriate in the common case, which is real-time 675 applications that do not have any CPU-bound threads. 676 677 Some real-time applications might not have a single real-time 678 thread that saturates a given CPU, but instead might have 679 multiple real-time threads that, taken together, fully utilize 680 that CPU. In this case, you should set RCU_BOOST_PRIO to 681 a priority higher than the lowest-priority thread that is 682 conspiring to prevent the CPU from running any non-real-time 683 tasks. For example, if one thread at priority 10 and another 684 thread at priority 5 are between themselves fully consuming 685 the CPU time on a given CPU, then RCU_BOOST_PRIO should be 686 set to priority 6 or higher. 687 688 Specify the real-time priority, or take the default if unsure. 689 690config RCU_BOOST_DELAY 691 int "Milliseconds to delay boosting after RCU grace-period start" 692 range 0 3000 693 depends on RCU_BOOST 694 default 500 695 help 696 This option specifies the time to wait after the beginning of 697 a given grace period before priority-boosting preempted RCU 698 readers blocking that grace period. Note that any RCU reader 699 blocking an expedited RCU grace period is boosted immediately. 700 701 Accept the default if unsure. 702 703config RCU_NOCB_CPU 704 bool "Offload RCU callback processing from boot-selected CPUs" 705 depends on TREE_RCU || TREE_PREEMPT_RCU 706 default n 707 help 708 Use this option to reduce OS jitter for aggressive HPC or 709 real-time workloads. It can also be used to offload RCU 710 callback invocation to energy-efficient CPUs in battery-powered 711 asymmetric multiprocessors. 712 713 This option offloads callback invocation from the set of 714 CPUs specified at boot time by the rcu_nocbs parameter. 715 For each such CPU, a kthread ("rcuox/N") will be created to 716 invoke callbacks, where the "N" is the CPU being offloaded, 717 and where the "x" is "b" for RCU-bh, "p" for RCU-preempt, and 718 "s" for RCU-sched. Nothing prevents this kthread from running 719 on the specified CPUs, but (1) the kthreads may be preempted 720 between each callback, and (2) affinity or cgroups can be used 721 to force the kthreads to run on whatever set of CPUs is desired. 722 723 Say Y here if you want to help to debug reduced OS jitter. 724 Say N here if you are unsure. 725 726choice 727 prompt "Build-forced no-CBs CPUs" 728 default RCU_NOCB_CPU_NONE 729 help 730 This option allows no-CBs CPUs (whose RCU callbacks are invoked 731 from kthreads rather than from softirq context) to be specified 732 at build time. Additional no-CBs CPUs may be specified by 733 the rcu_nocbs= boot parameter. 734 735config RCU_NOCB_CPU_NONE 736 bool "No build_forced no-CBs CPUs" 737 depends on RCU_NOCB_CPU && !NO_HZ_FULL 738 help 739 This option does not force any of the CPUs to be no-CBs CPUs. 740 Only CPUs designated by the rcu_nocbs= boot parameter will be 741 no-CBs CPUs, whose RCU callbacks will be invoked by per-CPU 742 kthreads whose names begin with "rcuo". All other CPUs will 743 invoke their own RCU callbacks in softirq context. 744 745 Select this option if you want to choose no-CBs CPUs at 746 boot time, for example, to allow testing of different no-CBs 747 configurations without having to rebuild the kernel each time. 748 749config RCU_NOCB_CPU_ZERO 750 bool "CPU 0 is a build_forced no-CBs CPU" 751 depends on RCU_NOCB_CPU && !NO_HZ_FULL 752 help 753 This option forces CPU 0 to be a no-CBs CPU, so that its RCU 754 callbacks are invoked by a per-CPU kthread whose name begins 755 with "rcuo". Additional CPUs may be designated as no-CBs 756 CPUs using the rcu_nocbs= boot parameter will be no-CBs CPUs. 757 All other CPUs will invoke their own RCU callbacks in softirq 758 context. 759 760 Select this if CPU 0 needs to be a no-CBs CPU for real-time 761 or energy-efficiency reasons, but the real reason it exists 762 is to ensure that randconfig testing covers mixed systems. 763 764config RCU_NOCB_CPU_ALL 765 bool "All CPUs are build_forced no-CBs CPUs" 766 depends on RCU_NOCB_CPU 767 help 768 This option forces all CPUs to be no-CBs CPUs. The rcu_nocbs= 769 boot parameter will be ignored. All CPUs' RCU callbacks will 770 be executed in the context of per-CPU rcuo kthreads created for 771 this purpose. Assuming that the kthreads whose names start with 772 "rcuo" are bound to "housekeeping" CPUs, this reduces OS jitter 773 on the remaining CPUs, but might decrease memory locality during 774 RCU-callback invocation, thus potentially degrading throughput. 775 776 Select this if all CPUs need to be no-CBs CPUs for real-time 777 or energy-efficiency reasons. 778 779endchoice 780 781endmenu # "RCU Subsystem" 782 783config IKCONFIG 784 tristate "Kernel .config support" 785 ---help--- 786 This option enables the complete Linux kernel ".config" file 787 contents to be saved in the kernel. It provides documentation 788 of which kernel options are used in a running kernel or in an 789 on-disk kernel. This information can be extracted from the kernel 790 image file with the script scripts/extract-ikconfig and used as 791 input to rebuild the current kernel or to build another kernel. 792 It can also be extracted from a running kernel by reading 793 /proc/config.gz if enabled (below). 794 795config IKCONFIG_PROC 796 bool "Enable access to .config through /proc/config.gz" 797 depends on IKCONFIG && PROC_FS 798 ---help--- 799 This option enables access to the kernel configuration file 800 through /proc/config.gz. 801 802config LOG_BUF_SHIFT 803 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 804 range 12 21 805 default 17 806 help 807 Select kernel log buffer size as a power of 2. 808 Examples: 809 17 => 128 KB 810 16 => 64 KB 811 15 => 32 KB 812 14 => 16 KB 813 13 => 8 KB 814 12 => 4 KB 815 816# 817# Architectures with an unreliable sched_clock() should select this: 818# 819config HAVE_UNSTABLE_SCHED_CLOCK 820 bool 821 822config GENERIC_SCHED_CLOCK 823 bool 824 825# 826# For architectures that want to enable the support for NUMA-affine scheduler 827# balancing logic: 828# 829config ARCH_SUPPORTS_NUMA_BALANCING 830 bool 831 832# For architectures that (ab)use NUMA to represent different memory regions 833# all cpu-local but of different latencies, such as SuperH. 834# 835config ARCH_WANT_NUMA_VARIABLE_LOCALITY 836 bool 837 838# 839# For architectures that are willing to define _PAGE_NUMA as _PAGE_PROTNONE 840config ARCH_WANTS_PROT_NUMA_PROT_NONE 841 bool 842 843config ARCH_USES_NUMA_PROT_NONE 844 bool 845 default y 846 depends on ARCH_WANTS_PROT_NUMA_PROT_NONE 847 depends on NUMA_BALANCING 848 849config NUMA_BALANCING_DEFAULT_ENABLED 850 bool "Automatically enable NUMA aware memory/task placement" 851 default y 852 depends on NUMA_BALANCING 853 help 854 If set, autonumic NUMA balancing will be enabled if running on a NUMA 855 machine. 856 857config NUMA_BALANCING 858 bool "Memory placement aware NUMA scheduler" 859 depends on ARCH_SUPPORTS_NUMA_BALANCING 860 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 861 depends on SMP && NUMA && MIGRATION 862 help 863 This option adds support for automatic NUMA aware memory/task placement. 864 The mechanism is quite primitive and is based on migrating memory when 865 it is references to the node the task is running on. 866 867 This system will be inactive on UMA systems. 868 869menuconfig CGROUPS 870 boolean "Control Group support" 871 depends on EVENTFD 872 help 873 This option adds support for grouping sets of processes together, for 874 use with process control subsystems such as Cpusets, CFS, memory 875 controls or device isolation. 876 See 877 - Documentation/scheduler/sched-design-CFS.txt (CFS) 878 - Documentation/cgroups/ (features for grouping, isolation 879 and resource control) 880 881 Say N if unsure. 882 883if CGROUPS 884 885config CGROUP_DEBUG 886 bool "Example debug cgroup subsystem" 887 default n 888 help 889 This option enables a simple cgroup subsystem that 890 exports useful debugging information about the cgroups 891 framework. 892 893 Say N if unsure. 894 895config CGROUP_FREEZER 896 bool "Freezer cgroup subsystem" 897 help 898 Provides a way to freeze and unfreeze all tasks in a 899 cgroup. 900 901config CGROUP_DEVICE 902 bool "Device controller for cgroups" 903 help 904 Provides a cgroup implementing whitelists for devices which 905 a process in the cgroup can mknod or open. 906 907config CPUSETS 908 bool "Cpuset support" 909 help 910 This option will let you create and manage CPUSETs which 911 allow dynamically partitioning a system into sets of CPUs and 912 Memory Nodes and assigning tasks to run only within those sets. 913 This is primarily useful on large SMP or NUMA systems. 914 915 Say N if unsure. 916 917config PROC_PID_CPUSET 918 bool "Include legacy /proc/<pid>/cpuset file" 919 depends on CPUSETS 920 default y 921 922config CGROUP_CPUACCT 923 bool "Simple CPU accounting cgroup subsystem" 924 help 925 Provides a simple Resource Controller for monitoring the 926 total CPU consumed by the tasks in a cgroup. 927 928config RESOURCE_COUNTERS 929 bool "Resource counters" 930 help 931 This option enables controller independent resource accounting 932 infrastructure that works with cgroups. 933 934config MEMCG 935 bool "Memory Resource Controller for Control Groups" 936 depends on RESOURCE_COUNTERS 937 select MM_OWNER 938 help 939 Provides a memory resource controller that manages both anonymous 940 memory and page cache. (See Documentation/cgroups/memory.txt) 941 942 Note that setting this option increases fixed memory overhead 943 associated with each page of memory in the system. By this, 944 8(16)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory 945 usage tracking struct at boot. Total amount of this is printed out 946 at boot. 947 948 Only enable when you're ok with these trade offs and really 949 sure you need the memory resource controller. Even when you enable 950 this, you can set "cgroup_disable=memory" at your boot option to 951 disable memory resource controller and you can avoid overheads. 952 (and lose benefits of memory resource controller) 953 954 This config option also selects MM_OWNER config option, which 955 could in turn add some fork/exit overhead. 956 957config MEMCG_SWAP 958 bool "Memory Resource Controller Swap Extension" 959 depends on MEMCG && SWAP 960 help 961 Add swap management feature to memory resource controller. When you 962 enable this, you can limit mem+swap usage per cgroup. In other words, 963 when you disable this, memory resource controller has no cares to 964 usage of swap...a process can exhaust all of the swap. This extension 965 is useful when you want to avoid exhaustion swap but this itself 966 adds more overheads and consumes memory for remembering information. 967 Especially if you use 32bit system or small memory system, please 968 be careful about enabling this. When memory resource controller 969 is disabled by boot option, this will be automatically disabled and 970 there will be no overhead from this. Even when you set this config=y, 971 if boot option "swapaccount=0" is set, swap will not be accounted. 972 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page 973 size is 4096bytes, 512k per 1Gbytes of swap. 974config MEMCG_SWAP_ENABLED 975 bool "Memory Resource Controller Swap Extension enabled by default" 976 depends on MEMCG_SWAP 977 default y 978 help 979 Memory Resource Controller Swap Extension comes with its price in 980 a bigger memory consumption. General purpose distribution kernels 981 which want to enable the feature but keep it disabled by default 982 and let the user enable it by swapaccount=1 boot command line 983 parameter should have this option unselected. 984 For those who want to have the feature enabled by default should 985 select this option (if, for some reason, they need to disable it 986 then swapaccount=0 does the trick). 987config MEMCG_KMEM 988 bool "Memory Resource Controller Kernel Memory accounting" 989 depends on MEMCG 990 depends on SLUB || SLAB 991 help 992 The Kernel Memory extension for Memory Resource Controller can limit 993 the amount of memory used by kernel objects in the system. Those are 994 fundamentally different from the entities handled by the standard 995 Memory Controller, which are page-based, and can be swapped. Users of 996 the kmem extension can use it to guarantee that no group of processes 997 will ever exhaust kernel resources alone. 998 999config CGROUP_HUGETLB 1000 bool "HugeTLB Resource Controller for Control Groups" 1001 depends on RESOURCE_COUNTERS && HUGETLB_PAGE 1002 default n 1003 help 1004 Provides a cgroup Resource Controller for HugeTLB pages. 1005 When you enable this, you can put a per cgroup limit on HugeTLB usage. 1006 The limit is enforced during page fault. Since HugeTLB doesn't 1007 support page reclaim, enforcing the limit at page fault time implies 1008 that, the application will get SIGBUS signal if it tries to access 1009 HugeTLB pages beyond its limit. This requires the application to know 1010 beforehand how much HugeTLB pages it would require for its use. The 1011 control group is tracked in the third page lru pointer. This means 1012 that we cannot use the controller with huge page less than 3 pages. 1013 1014config CGROUP_PERF 1015 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring" 1016 depends on PERF_EVENTS && CGROUPS 1017 help 1018 This option extends the per-cpu mode to restrict monitoring to 1019 threads which belong to the cgroup specified and run on the 1020 designated cpu. 1021 1022 Say N if unsure. 1023 1024menuconfig CGROUP_SCHED 1025 bool "Group CPU scheduler" 1026 default n 1027 help 1028 This feature lets CPU scheduler recognize task groups and control CPU 1029 bandwidth allocation to such task groups. It uses cgroups to group 1030 tasks. 1031 1032if CGROUP_SCHED 1033config FAIR_GROUP_SCHED 1034 bool "Group scheduling for SCHED_OTHER" 1035 depends on CGROUP_SCHED 1036 default CGROUP_SCHED 1037 1038config CFS_BANDWIDTH 1039 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 1040 depends on FAIR_GROUP_SCHED 1041 default n 1042 help 1043 This option allows users to define CPU bandwidth rates (limits) for 1044 tasks running within the fair group scheduler. Groups with no limit 1045 set are considered to be unconstrained and will run with no 1046 restriction. 1047 See tip/Documentation/scheduler/sched-bwc.txt for more information. 1048 1049config RT_GROUP_SCHED 1050 bool "Group scheduling for SCHED_RR/FIFO" 1051 depends on CGROUP_SCHED 1052 default n 1053 help 1054 This feature lets you explicitly allocate real CPU bandwidth 1055 to task groups. If enabled, it will also make it impossible to 1056 schedule realtime tasks for non-root users until you allocate 1057 realtime bandwidth for them. 1058 See Documentation/scheduler/sched-rt-group.txt for more information. 1059 1060endif #CGROUP_SCHED 1061 1062config BLK_CGROUP 1063 bool "Block IO controller" 1064 depends on BLOCK 1065 default n 1066 ---help--- 1067 Generic block IO controller cgroup interface. This is the common 1068 cgroup interface which should be used by various IO controlling 1069 policies. 1070 1071 Currently, CFQ IO scheduler uses it to recognize task groups and 1072 control disk bandwidth allocation (proportional time slice allocation) 1073 to such task groups. It is also used by bio throttling logic in 1074 block layer to implement upper limit in IO rates on a device. 1075 1076 This option only enables generic Block IO controller infrastructure. 1077 One needs to also enable actual IO controlling logic/policy. For 1078 enabling proportional weight division of disk bandwidth in CFQ, set 1079 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 1080 CONFIG_BLK_DEV_THROTTLING=y. 1081 1082 See Documentation/cgroups/blkio-controller.txt for more information. 1083 1084config DEBUG_BLK_CGROUP 1085 bool "Enable Block IO controller debugging" 1086 depends on BLK_CGROUP 1087 default n 1088 ---help--- 1089 Enable some debugging help. Currently it exports additional stat 1090 files in a cgroup which can be useful for debugging. 1091 1092endif # CGROUPS 1093 1094config CHECKPOINT_RESTORE 1095 bool "Checkpoint/restore support" if EXPERT 1096 default n 1097 help 1098 Enables additional kernel features in a sake of checkpoint/restore. 1099 In particular it adds auxiliary prctl codes to setup process text, 1100 data and heap segment sizes, and a few additional /proc filesystem 1101 entries. 1102 1103 If unsure, say N here. 1104 1105menuconfig NAMESPACES 1106 bool "Namespaces support" if EXPERT 1107 default !EXPERT 1108 help 1109 Provides the way to make tasks work with different objects using 1110 the same id. For example same IPC id may refer to different objects 1111 or same user id or pid may refer to different tasks when used in 1112 different namespaces. 1113 1114if NAMESPACES 1115 1116config UTS_NS 1117 bool "UTS namespace" 1118 default y 1119 help 1120 In this namespace tasks see different info provided with the 1121 uname() system call 1122 1123config IPC_NS 1124 bool "IPC namespace" 1125 depends on (SYSVIPC || POSIX_MQUEUE) 1126 default y 1127 help 1128 In this namespace tasks work with IPC ids which correspond to 1129 different IPC objects in different namespaces. 1130 1131config USER_NS 1132 bool "User namespace" 1133 select UIDGID_STRICT_TYPE_CHECKS 1134 1135 default n 1136 help 1137 This allows containers, i.e. vservers, to use user namespaces 1138 to provide different user info for different servers. 1139 1140 When user namespaces are enabled in the kernel it is 1141 recommended that the MEMCG and MEMCG_KMEM options also be 1142 enabled and that user-space use the memory control groups to 1143 limit the amount of memory a memory unprivileged users can 1144 use. 1145 1146 If unsure, say N. 1147 1148config PID_NS 1149 bool "PID Namespaces" 1150 default y 1151 help 1152 Support process id namespaces. This allows having multiple 1153 processes with the same pid as long as they are in different 1154 pid namespaces. This is a building block of containers. 1155 1156config NET_NS 1157 bool "Network namespace" 1158 depends on NET 1159 default y 1160 help 1161 Allow user space to create what appear to be multiple instances 1162 of the network stack. 1163 1164endif # NAMESPACES 1165 1166config UIDGID_STRICT_TYPE_CHECKS 1167 bool "Require conversions between uid/gids and their internal representation" 1168 default n 1169 help 1170 While the nececessary conversions are being added to all subsystems this option allows 1171 the code to continue to build for unconverted subsystems. 1172 1173 Say Y here if you want the strict type checking enabled 1174 1175config SCHED_AUTOGROUP 1176 bool "Automatic process group scheduling" 1177 select EVENTFD 1178 select CGROUPS 1179 select CGROUP_SCHED 1180 select FAIR_GROUP_SCHED 1181 help 1182 This option optimizes the scheduler for common desktop workloads by 1183 automatically creating and populating task groups. This separation 1184 of workloads isolates aggressive CPU burners (like build jobs) from 1185 desktop applications. Task group autogeneration is currently based 1186 upon task session. 1187 1188config MM_OWNER 1189 bool 1190 1191config SYSFS_DEPRECATED 1192 bool "Enable deprecated sysfs features to support old userspace tools" 1193 depends on SYSFS 1194 default n 1195 help 1196 This option adds code that switches the layout of the "block" class 1197 devices, to not show up in /sys/class/block/, but only in 1198 /sys/block/. 1199 1200 This switch is only active when the sysfs.deprecated=1 boot option is 1201 passed or the SYSFS_DEPRECATED_V2 option is set. 1202 1203 This option allows new kernels to run on old distributions and tools, 1204 which might get confused by /sys/class/block/. Since 2007/2008 all 1205 major distributions and tools handle this just fine. 1206 1207 Recent distributions and userspace tools after 2009/2010 depend on 1208 the existence of /sys/class/block/, and will not work with this 1209 option enabled. 1210 1211 Only if you are using a new kernel on an old distribution, you might 1212 need to say Y here. 1213 1214config SYSFS_DEPRECATED_V2 1215 bool "Enable deprecated sysfs features by default" 1216 default n 1217 depends on SYSFS 1218 depends on SYSFS_DEPRECATED 1219 help 1220 Enable deprecated sysfs by default. 1221 1222 See the CONFIG_SYSFS_DEPRECATED option for more details about this 1223 option. 1224 1225 Only if you are using a new kernel on an old distribution, you might 1226 need to say Y here. Even then, odds are you would not need it 1227 enabled, you can always pass the boot option if absolutely necessary. 1228 1229config RELAY 1230 bool "Kernel->user space relay support (formerly relayfs)" 1231 help 1232 This option enables support for relay interface support in 1233 certain file systems (such as debugfs). 1234 It is designed to provide an efficient mechanism for tools and 1235 facilities to relay large amounts of data from kernel space to 1236 user space. 1237 1238 If unsure, say N. 1239 1240config BLK_DEV_INITRD 1241 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1242 depends on BROKEN || !FRV 1243 help 1244 The initial RAM filesystem is a ramfs which is loaded by the 1245 boot loader (loadlin or lilo) and that is mounted as root 1246 before the normal boot procedure. It is typically used to 1247 load modules needed to mount the "real" root file system, 1248 etc. See <file:Documentation/initrd.txt> for details. 1249 1250 If RAM disk support (BLK_DEV_RAM) is also included, this 1251 also enables initial RAM disk (initrd) support and adds 1252 15 Kbytes (more on some other architectures) to the kernel size. 1253 1254 If unsure say Y. 1255 1256if BLK_DEV_INITRD 1257 1258source "usr/Kconfig" 1259 1260endif 1261 1262config CC_OPTIMIZE_FOR_SIZE 1263 bool "Optimize for size" 1264 help 1265 Enabling this option will pass "-Os" instead of "-O2" to gcc 1266 resulting in a smaller kernel. 1267 1268 If unsure, say N. 1269 1270config SYSCTL 1271 bool 1272 1273config ANON_INODES 1274 bool 1275 1276config HAVE_UID16 1277 bool 1278 1279config SYSCTL_EXCEPTION_TRACE 1280 bool 1281 help 1282 Enable support for /proc/sys/debug/exception-trace. 1283 1284config SYSCTL_ARCH_UNALIGN_NO_WARN 1285 bool 1286 help 1287 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1288 Allows arch to define/use @no_unaligned_warning to possibly warn 1289 about unaligned access emulation going on under the hood. 1290 1291config SYSCTL_ARCH_UNALIGN_ALLOW 1292 bool 1293 help 1294 Enable support for /proc/sys/kernel/unaligned-trap 1295 Allows arches to define/use @unaligned_enabled to runtime toggle 1296 the unaligned access emulation. 1297 see arch/parisc/kernel/unaligned.c for reference 1298 1299config HAVE_PCSPKR_PLATFORM 1300 bool 1301 1302menuconfig EXPERT 1303 bool "Configure standard kernel features (expert users)" 1304 # Unhide debug options, to make the on-by-default options visible 1305 select DEBUG_KERNEL 1306 help 1307 This option allows certain base kernel options and settings 1308 to be disabled or tweaked. This is for specialized 1309 environments which can tolerate a "non-standard" kernel. 1310 Only use this if you really know what you are doing. 1311 1312config UID16 1313 bool "Enable 16-bit UID system calls" if EXPERT 1314 depends on HAVE_UID16 1315 default y 1316 help 1317 This enables the legacy 16-bit UID syscall wrappers. 1318 1319config SYSCTL_SYSCALL 1320 bool "Sysctl syscall support" if EXPERT 1321 depends on PROC_SYSCTL 1322 default n 1323 select SYSCTL 1324 ---help--- 1325 sys_sysctl uses binary paths that have been found challenging 1326 to properly maintain and use. The interface in /proc/sys 1327 using paths with ascii names is now the primary path to this 1328 information. 1329 1330 Almost nothing using the binary sysctl interface so if you are 1331 trying to save some space it is probably safe to disable this, 1332 making your kernel marginally smaller. 1333 1334 If unsure say N here. 1335 1336config KALLSYMS 1337 bool "Load all symbols for debugging/ksymoops" if EXPERT 1338 default y 1339 help 1340 Say Y here to let the kernel print out symbolic crash information and 1341 symbolic stack backtraces. This increases the size of the kernel 1342 somewhat, as all symbols have to be loaded into the kernel image. 1343 1344config KALLSYMS_ALL 1345 bool "Include all symbols in kallsyms" 1346 depends on DEBUG_KERNEL && KALLSYMS 1347 help 1348 Normally kallsyms only contains the symbols of functions for nicer 1349 OOPS messages and backtraces (i.e., symbols from the text and inittext 1350 sections). This is sufficient for most cases. And only in very rare 1351 cases (e.g., when a debugger is used) all symbols are required (e.g., 1352 names of variables from the data sections, etc). 1353 1354 This option makes sure that all symbols are loaded into the kernel 1355 image (i.e., symbols from all sections) in cost of increased kernel 1356 size (depending on the kernel configuration, it may be 300KiB or 1357 something like this). 1358 1359 Say N unless you really need all symbols. 1360 1361config PRINTK 1362 default y 1363 bool "Enable support for printk" if EXPERT 1364 select IRQ_WORK 1365 help 1366 This option enables normal printk support. Removing it 1367 eliminates most of the message strings from the kernel image 1368 and makes the kernel more or less silent. As this makes it 1369 very difficult to diagnose system problems, saying N here is 1370 strongly discouraged. 1371 1372config BUG 1373 bool "BUG() support" if EXPERT 1374 default y 1375 help 1376 Disabling this option eliminates support for BUG and WARN, reducing 1377 the size of your kernel image and potentially quietly ignoring 1378 numerous fatal conditions. You should only consider disabling this 1379 option for embedded systems with no facilities for reporting errors. 1380 Just say Y. 1381 1382config ELF_CORE 1383 depends on COREDUMP 1384 default y 1385 bool "Enable ELF core dumps" if EXPERT 1386 help 1387 Enable support for generating core dumps. Disabling saves about 4k. 1388 1389 1390config PCSPKR_PLATFORM 1391 bool "Enable PC-Speaker support" if EXPERT 1392 depends on HAVE_PCSPKR_PLATFORM 1393 select I8253_LOCK 1394 default y 1395 help 1396 This option allows to disable the internal PC-Speaker 1397 support, saving some memory. 1398 1399config BASE_FULL 1400 default y 1401 bool "Enable full-sized data structures for core" if EXPERT 1402 help 1403 Disabling this option reduces the size of miscellaneous core 1404 kernel data structures. This saves memory on small machines, 1405 but may reduce performance. 1406 1407config FUTEX 1408 bool "Enable futex support" if EXPERT 1409 default y 1410 select RT_MUTEXES 1411 help 1412 Disabling this option will cause the kernel to be built without 1413 support for "fast userspace mutexes". The resulting kernel may not 1414 run glibc-based applications correctly. 1415 1416config EPOLL 1417 bool "Enable eventpoll support" if EXPERT 1418 default y 1419 select ANON_INODES 1420 help 1421 Disabling this option will cause the kernel to be built without 1422 support for epoll family of system calls. 1423 1424config SIGNALFD 1425 bool "Enable signalfd() system call" if EXPERT 1426 select ANON_INODES 1427 default y 1428 help 1429 Enable the signalfd() system call that allows to receive signals 1430 on a file descriptor. 1431 1432 If unsure, say Y. 1433 1434config TIMERFD 1435 bool "Enable timerfd() system call" if EXPERT 1436 select ANON_INODES 1437 default y 1438 help 1439 Enable the timerfd() system call that allows to receive timer 1440 events on a file descriptor. 1441 1442 If unsure, say Y. 1443 1444config EVENTFD 1445 bool "Enable eventfd() system call" if EXPERT 1446 select ANON_INODES 1447 default y 1448 help 1449 Enable the eventfd() system call that allows to receive both 1450 kernel notification (ie. KAIO) or userspace notifications. 1451 1452 If unsure, say Y. 1453 1454config SHMEM 1455 bool "Use full shmem filesystem" if EXPERT 1456 default y 1457 depends on MMU 1458 help 1459 The shmem is an internal filesystem used to manage shared memory. 1460 It is backed by swap and manages resource limits. It is also exported 1461 to userspace as tmpfs if TMPFS is enabled. Disabling this 1462 option replaces shmem and tmpfs with the much simpler ramfs code, 1463 which may be appropriate on small systems without swap. 1464 1465config AIO 1466 bool "Enable AIO support" if EXPERT 1467 default y 1468 help 1469 This option enables POSIX asynchronous I/O which may by used 1470 by some high performance threaded applications. Disabling 1471 this option saves about 7k. 1472 1473config PCI_QUIRKS 1474 default y 1475 bool "Enable PCI quirk workarounds" if EXPERT 1476 depends on PCI 1477 help 1478 This enables workarounds for various PCI chipset 1479 bugs/quirks. Disable this only if your target machine is 1480 unaffected by PCI quirks. 1481 1482config EMBEDDED 1483 bool "Embedded system" 1484 select EXPERT 1485 help 1486 This option should be enabled if compiling the kernel for 1487 an embedded system so certain expert options are available 1488 for configuration. 1489 1490config HAVE_PERF_EVENTS 1491 bool 1492 help 1493 See tools/perf/design.txt for details. 1494 1495config PERF_USE_VMALLOC 1496 bool 1497 help 1498 See tools/perf/design.txt for details 1499 1500menu "Kernel Performance Events And Counters" 1501 1502config PERF_EVENTS 1503 bool "Kernel performance events and counters" 1504 default y if PROFILING 1505 depends on HAVE_PERF_EVENTS 1506 select ANON_INODES 1507 select IRQ_WORK 1508 help 1509 Enable kernel support for various performance events provided 1510 by software and hardware. 1511 1512 Software events are supported either built-in or via the 1513 use of generic tracepoints. 1514 1515 Most modern CPUs support performance events via performance 1516 counter registers. These registers count the number of certain 1517 types of hw events: such as instructions executed, cachemisses 1518 suffered, or branches mis-predicted - without slowing down the 1519 kernel or applications. These registers can also trigger interrupts 1520 when a threshold number of events have passed - and can thus be 1521 used to profile the code that runs on that CPU. 1522 1523 The Linux Performance Event subsystem provides an abstraction of 1524 these software and hardware event capabilities, available via a 1525 system call and used by the "perf" utility in tools/perf/. It 1526 provides per task and per CPU counters, and it provides event 1527 capabilities on top of those. 1528 1529 Say Y if unsure. 1530 1531config DEBUG_PERF_USE_VMALLOC 1532 default n 1533 bool "Debug: use vmalloc to back perf mmap() buffers" 1534 depends on PERF_EVENTS && DEBUG_KERNEL 1535 select PERF_USE_VMALLOC 1536 help 1537 Use vmalloc memory to back perf mmap() buffers. 1538 1539 Mostly useful for debugging the vmalloc code on platforms 1540 that don't require it. 1541 1542 Say N if unsure. 1543 1544endmenu 1545 1546config VM_EVENT_COUNTERS 1547 default y 1548 bool "Enable VM event counters for /proc/vmstat" if EXPERT 1549 help 1550 VM event counters are needed for event counts to be shown. 1551 This option allows the disabling of the VM event counters 1552 on EXPERT systems. /proc/vmstat will only show page counts 1553 if VM event counters are disabled. 1554 1555config SLUB_DEBUG 1556 default y 1557 bool "Enable SLUB debugging support" if EXPERT 1558 depends on SLUB && SYSFS 1559 help 1560 SLUB has extensive debug support features. Disabling these can 1561 result in significant savings in code size. This also disables 1562 SLUB sysfs support. /sys/slab will not exist and there will be 1563 no support for cache validation etc. 1564 1565config COMPAT_BRK 1566 bool "Disable heap randomization" 1567 default y 1568 help 1569 Randomizing heap placement makes heap exploits harder, but it 1570 also breaks ancient binaries (including anything libc5 based). 1571 This option changes the bootup default to heap randomization 1572 disabled, and can be overridden at runtime by setting 1573 /proc/sys/kernel/randomize_va_space to 2. 1574 1575 On non-ancient distros (post-2000 ones) N is usually a safe choice. 1576 1577choice 1578 prompt "Choose SLAB allocator" 1579 default SLUB 1580 help 1581 This option allows to select a slab allocator. 1582 1583config SLAB 1584 bool "SLAB" 1585 help 1586 The regular slab allocator that is established and known to work 1587 well in all environments. It organizes cache hot objects in 1588 per cpu and per node queues. 1589 1590config SLUB 1591 bool "SLUB (Unqueued Allocator)" 1592 help 1593 SLUB is a slab allocator that minimizes cache line usage 1594 instead of managing queues of cached objects (SLAB approach). 1595 Per cpu caching is realized using slabs of objects instead 1596 of queues of objects. SLUB can use memory efficiently 1597 and has enhanced diagnostics. SLUB is the default choice for 1598 a slab allocator. 1599 1600config SLOB 1601 depends on EXPERT 1602 bool "SLOB (Simple Allocator)" 1603 help 1604 SLOB replaces the stock allocator with a drastically simpler 1605 allocator. SLOB is generally more space efficient but 1606 does not perform as well on large systems. 1607 1608endchoice 1609 1610config SLUB_CPU_PARTIAL 1611 default y 1612 depends on SLUB && SMP 1613 bool "SLUB per cpu partial cache" 1614 help 1615 Per cpu partial caches accellerate objects allocation and freeing 1616 that is local to a processor at the price of more indeterminism 1617 in the latency of the free. On overflow these caches will be cleared 1618 which requires the taking of locks that may cause latency spikes. 1619 Typically one would choose no for a realtime system. 1620 1621config MMAP_ALLOW_UNINITIALIZED 1622 bool "Allow mmapped anonymous memory to be uninitialized" 1623 depends on EXPERT && !MMU 1624 default n 1625 help 1626 Normally, and according to the Linux spec, anonymous memory obtained 1627 from mmap() has it's contents cleared before it is passed to 1628 userspace. Enabling this config option allows you to request that 1629 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1630 providing a huge performance boost. If this option is not enabled, 1631 then the flag will be ignored. 1632 1633 This is taken advantage of by uClibc's malloc(), and also by 1634 ELF-FDPIC binfmt's brk and stack allocator. 1635 1636 Because of the obvious security issues, this option should only be 1637 enabled on embedded devices where you control what is run in 1638 userspace. Since that isn't generally a problem on no-MMU systems, 1639 it is normally safe to say Y here. 1640 1641 See Documentation/nommu-mmap.txt for more information. 1642 1643config PROFILING 1644 bool "Profiling support" 1645 help 1646 Say Y here to enable the extended profiling support mechanisms used 1647 by profilers such as OProfile. 1648 1649# 1650# Place an empty function call at each tracepoint site. Can be 1651# dynamically changed for a probe function. 1652# 1653config TRACEPOINTS 1654 bool 1655 1656source "arch/Kconfig" 1657 1658endmenu # General setup 1659 1660config HAVE_GENERIC_DMA_COHERENT 1661 bool 1662 default n 1663 1664config SLABINFO 1665 bool 1666 depends on PROC_FS 1667 depends on SLAB || SLUB_DEBUG 1668 default y 1669 1670config RT_MUTEXES 1671 boolean 1672 1673config BASE_SMALL 1674 int 1675 default 0 if BASE_FULL 1676 default 1 if !BASE_FULL 1677 1678menuconfig MODULES 1679 bool "Enable loadable module support" 1680 option modules 1681 help 1682 Kernel modules are small pieces of compiled code which can 1683 be inserted in the running kernel, rather than being 1684 permanently built into the kernel. You use the "modprobe" 1685 tool to add (and sometimes remove) them. If you say Y here, 1686 many parts of the kernel can be built as modules (by 1687 answering M instead of Y where indicated): this is most 1688 useful for infrequently used options which are not required 1689 for booting. For more information, see the man pages for 1690 modprobe, lsmod, modinfo, insmod and rmmod. 1691 1692 If you say Y here, you will need to run "make 1693 modules_install" to put the modules under /lib/modules/ 1694 where modprobe can find them (you may need to be root to do 1695 this). 1696 1697 If unsure, say Y. 1698 1699if MODULES 1700 1701config MODULE_FORCE_LOAD 1702 bool "Forced module loading" 1703 default n 1704 help 1705 Allow loading of modules without version information (ie. modprobe 1706 --force). Forced module loading sets the 'F' (forced) taint flag and 1707 is usually a really bad idea. 1708 1709config MODULE_UNLOAD 1710 bool "Module unloading" 1711 help 1712 Without this option you will not be able to unload any 1713 modules (note that some modules may not be unloadable 1714 anyway), which makes your kernel smaller, faster 1715 and simpler. If unsure, say Y. 1716 1717config MODULE_FORCE_UNLOAD 1718 bool "Forced module unloading" 1719 depends on MODULE_UNLOAD 1720 help 1721 This option allows you to force a module to unload, even if the 1722 kernel believes it is unsafe: the kernel will remove the module 1723 without waiting for anyone to stop using it (using the -f option to 1724 rmmod). This is mainly for kernel developers and desperate users. 1725 If unsure, say N. 1726 1727config MODVERSIONS 1728 bool "Module versioning support" 1729 help 1730 Usually, you have to use modules compiled with your kernel. 1731 Saying Y here makes it sometimes possible to use modules 1732 compiled for different kernels, by adding enough information 1733 to the modules to (hopefully) spot any changes which would 1734 make them incompatible with the kernel you are running. If 1735 unsure, say N. 1736 1737config MODULE_SRCVERSION_ALL 1738 bool "Source checksum for all modules" 1739 help 1740 Modules which contain a MODULE_VERSION get an extra "srcversion" 1741 field inserted into their modinfo section, which contains a 1742 sum of the source files which made it. This helps maintainers 1743 see exactly which source was used to build a module (since 1744 others sometimes change the module source without updating 1745 the version). With this option, such a "srcversion" field 1746 will be created for all modules. If unsure, say N. 1747 1748config MODULE_SIG 1749 bool "Module signature verification" 1750 depends on MODULES 1751 select KEYS 1752 select CRYPTO 1753 select ASYMMETRIC_KEY_TYPE 1754 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1755 select PUBLIC_KEY_ALGO_RSA 1756 select ASN1 1757 select OID_REGISTRY 1758 select X509_CERTIFICATE_PARSER 1759 help 1760 Check modules for valid signatures upon load: the signature 1761 is simply appended to the module. For more information see 1762 Documentation/module-signing.txt. 1763 1764 !!!WARNING!!! If you enable this option, you MUST make sure that the 1765 module DOES NOT get stripped after being signed. This includes the 1766 debuginfo strip done by some packagers (such as rpmbuild) and 1767 inclusion into an initramfs that wants the module size reduced. 1768 1769config MODULE_SIG_FORCE 1770 bool "Require modules to be validly signed" 1771 depends on MODULE_SIG 1772 help 1773 Reject unsigned modules or signed modules for which we don't have a 1774 key. Without this, such modules will simply taint the kernel. 1775 1776config MODULE_SIG_ALL 1777 bool "Automatically sign all modules" 1778 default y 1779 depends on MODULE_SIG 1780 help 1781 Sign all modules during make modules_install. Without this option, 1782 modules must be signed manually, using the scripts/sign-file tool. 1783 1784comment "Do not forget to sign required modules with scripts/sign-file" 1785 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 1786 1787choice 1788 prompt "Which hash algorithm should modules be signed with?" 1789 depends on MODULE_SIG 1790 help 1791 This determines which sort of hashing algorithm will be used during 1792 signature generation. This algorithm _must_ be built into the kernel 1793 directly so that signature verification can take place. It is not 1794 possible to load a signed module containing the algorithm to check 1795 the signature on that module. 1796 1797config MODULE_SIG_SHA1 1798 bool "Sign modules with SHA-1" 1799 select CRYPTO_SHA1 1800 1801config MODULE_SIG_SHA224 1802 bool "Sign modules with SHA-224" 1803 select CRYPTO_SHA256 1804 1805config MODULE_SIG_SHA256 1806 bool "Sign modules with SHA-256" 1807 select CRYPTO_SHA256 1808 1809config MODULE_SIG_SHA384 1810 bool "Sign modules with SHA-384" 1811 select CRYPTO_SHA512 1812 1813config MODULE_SIG_SHA512 1814 bool "Sign modules with SHA-512" 1815 select CRYPTO_SHA512 1816 1817endchoice 1818 1819config MODULE_SIG_HASH 1820 string 1821 depends on MODULE_SIG 1822 default "sha1" if MODULE_SIG_SHA1 1823 default "sha224" if MODULE_SIG_SHA224 1824 default "sha256" if MODULE_SIG_SHA256 1825 default "sha384" if MODULE_SIG_SHA384 1826 default "sha512" if MODULE_SIG_SHA512 1827 1828endif # MODULES 1829 1830config INIT_ALL_POSSIBLE 1831 bool 1832 help 1833 Back when each arch used to define their own cpu_online_mask and 1834 cpu_possible_mask, some of them chose to initialize cpu_possible_mask 1835 with all 1s, and others with all 0s. When they were centralised, 1836 it was better to provide this option than to break all the archs 1837 and have several arch maintainers pursuing me down dark alleys. 1838 1839config STOP_MACHINE 1840 bool 1841 default y 1842 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU 1843 help 1844 Need stop_machine() primitive. 1845 1846source "block/Kconfig" 1847 1848config PREEMPT_NOTIFIERS 1849 bool 1850 1851config PADATA 1852 depends on SMP 1853 bool 1854 1855# Can be selected by architectures with broken toolchains 1856# that get confused by correct const<->read_only section 1857# mappings 1858config BROKEN_RODATA 1859 bool 1860 1861config ASN1 1862 tristate 1863 help 1864 Build a simple ASN.1 grammar compiler that produces a bytecode output 1865 that can be interpreted by the ASN.1 stream decoder and used to 1866 inform it as to what tags are to be expected in a stream and what 1867 functions to call on what tags. 1868 1869source "kernel/Kconfig.locks" 1870