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 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 122 help 123 The linux kernel is a kind of self-extracting executable. 124 Several compression algorithms are available, which differ 125 in efficiency, compression and decompression speed. 126 Compression speed is only relevant when building a kernel. 127 Decompression speed is relevant at each boot. 128 129 If you have any problems with bzip2 or lzma compressed 130 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 131 version of this functionality (bzip2 only), for 2.4, was 132 supplied by Christian Ludwig) 133 134 High compression options are mostly useful for users, who 135 are low on disk space (embedded systems), but for whom ram 136 size matters less. 137 138 If in doubt, select 'gzip' 139 140config KERNEL_GZIP 141 bool "Gzip" 142 depends on HAVE_KERNEL_GZIP 143 help 144 The old and tried gzip compression. It provides a good balance 145 between compression ratio and decompression speed. 146 147config KERNEL_BZIP2 148 bool "Bzip2" 149 depends on HAVE_KERNEL_BZIP2 150 help 151 Its compression ratio and speed is intermediate. 152 Decompression speed is slowest among the choices. The kernel 153 size is about 10% smaller with bzip2, in comparison to gzip. 154 Bzip2 uses a large amount of memory. For modern kernels you 155 will need at least 8MB RAM or more for booting. 156 157config KERNEL_LZMA 158 bool "LZMA" 159 depends on HAVE_KERNEL_LZMA 160 help 161 This compression algorithm's ratio is best. Decompression speed 162 is between gzip and bzip2. Compression is slowest. 163 The kernel size is about 33% smaller with LZMA in comparison to gzip. 164 165config KERNEL_XZ 166 bool "XZ" 167 depends on HAVE_KERNEL_XZ 168 help 169 XZ uses the LZMA2 algorithm and instruction set specific 170 BCJ filters which can improve compression ratio of executable 171 code. The size of the kernel is about 30% smaller with XZ in 172 comparison to gzip. On architectures for which there is a BCJ 173 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 174 will create a few percent smaller kernel than plain LZMA. 175 176 The speed is about the same as with LZMA: The decompression 177 speed of XZ is better than that of bzip2 but worse than gzip 178 and LZO. Compression is slow. 179 180config KERNEL_LZO 181 bool "LZO" 182 depends on HAVE_KERNEL_LZO 183 help 184 Its compression ratio is the poorest among the choices. The kernel 185 size is about 10% bigger than gzip; however its speed 186 (both compression and decompression) is the fastest. 187 188config KERNEL_LZ4 189 bool "LZ4" 190 depends on HAVE_KERNEL_LZ4 191 help 192 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 193 A preliminary version of LZ4 de/compression tool is available at 194 <https://code.google.com/p/lz4/>. 195 196 Its compression ratio is worse than LZO. The size of the kernel 197 is about 8% bigger than LZO. But the decompression speed is 198 faster than LZO. 199 200endchoice 201 202config DEFAULT_HOSTNAME 203 string "Default hostname" 204 default "(none)" 205 help 206 This option determines the default system hostname before userspace 207 calls sethostname(2). The kernel traditionally uses "(none)" here, 208 but you may wish to use a different default here to make a minimal 209 system more usable with less configuration. 210 211config SWAP 212 bool "Support for paging of anonymous memory (swap)" 213 depends on MMU && BLOCK 214 default y 215 help 216 This option allows you to choose whether you want to have support 217 for so called swap devices or swap files in your kernel that are 218 used to provide more virtual memory than the actual RAM present 219 in your computer. If unsure say Y. 220 221config SYSVIPC 222 bool "System V IPC" 223 ---help--- 224 Inter Process Communication is a suite of library functions and 225 system calls which let processes (running programs) synchronize and 226 exchange information. It is generally considered to be a good thing, 227 and some programs won't run unless you say Y here. In particular, if 228 you want to run the DOS emulator dosemu under Linux (read the 229 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 230 you'll need to say Y here. 231 232 You can find documentation about IPC with "info ipc" and also in 233 section 6.4 of the Linux Programmer's Guide, available from 234 <http://www.tldp.org/guides.html>. 235 236config SYSVIPC_SYSCTL 237 bool 238 depends on SYSVIPC 239 depends on SYSCTL 240 default y 241 242config POSIX_MQUEUE 243 bool "POSIX Message Queues" 244 depends on NET 245 ---help--- 246 POSIX variant of message queues is a part of IPC. In POSIX message 247 queues every message has a priority which decides about succession 248 of receiving it by a process. If you want to compile and run 249 programs written e.g. for Solaris with use of its POSIX message 250 queues (functions mq_*) say Y here. 251 252 POSIX message queues are visible as a filesystem called 'mqueue' 253 and can be mounted somewhere if you want to do filesystem 254 operations on message queues. 255 256 If unsure, say Y. 257 258config POSIX_MQUEUE_SYSCTL 259 bool 260 depends on POSIX_MQUEUE 261 depends on SYSCTL 262 default y 263 264config FHANDLE 265 bool "open by fhandle syscalls" 266 select EXPORTFS 267 help 268 If you say Y here, a user level program will be able to map 269 file names to handle and then later use the handle for 270 different file system operations. This is useful in implementing 271 userspace file servers, which now track files using handles instead 272 of names. The handle would remain the same even if file names 273 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 274 syscalls. 275 276config AUDIT 277 bool "Auditing support" 278 depends on NET 279 help 280 Enable auditing infrastructure that can be used with another 281 kernel subsystem, such as SELinux (which requires this for 282 logging of avc messages output). Does not do system-call 283 auditing without CONFIG_AUDITSYSCALL. 284 285config AUDITSYSCALL 286 bool "Enable system-call auditing support" 287 depends on AUDIT && (X86 || PARISC || PPC || S390 || IA64 || UML || SPARC64 || SUPERH || (ARM && AEABI && !OABI_COMPAT)) 288 default y if SECURITY_SELINUX 289 help 290 Enable low-overhead system-call auditing infrastructure that 291 can be used independently or with another kernel subsystem, 292 such as SELinux. 293 294config AUDIT_WATCH 295 def_bool y 296 depends on AUDITSYSCALL 297 select FSNOTIFY 298 299config AUDIT_TREE 300 def_bool y 301 depends on AUDITSYSCALL 302 select FSNOTIFY 303 304source "kernel/irq/Kconfig" 305source "kernel/time/Kconfig" 306 307menu "CPU/Task time and stats accounting" 308 309config VIRT_CPU_ACCOUNTING 310 bool 311 312choice 313 prompt "Cputime accounting" 314 default TICK_CPU_ACCOUNTING if !PPC64 315 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 316 317# Kind of a stub config for the pure tick based cputime accounting 318config TICK_CPU_ACCOUNTING 319 bool "Simple tick based cputime accounting" 320 depends on !S390 && !NO_HZ_FULL 321 help 322 This is the basic tick based cputime accounting that maintains 323 statistics about user, system and idle time spent on per jiffies 324 granularity. 325 326 If unsure, say Y. 327 328config VIRT_CPU_ACCOUNTING_NATIVE 329 bool "Deterministic task and CPU time accounting" 330 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 331 select VIRT_CPU_ACCOUNTING 332 help 333 Select this option to enable more accurate task and CPU time 334 accounting. This is done by reading a CPU counter on each 335 kernel entry and exit and on transitions within the kernel 336 between system, softirq and hardirq state, so there is a 337 small performance impact. In the case of s390 or IBM POWER > 5, 338 this also enables accounting of stolen time on logically-partitioned 339 systems. 340 341config VIRT_CPU_ACCOUNTING_GEN 342 bool "Full dynticks CPU time accounting" 343 depends on HAVE_CONTEXT_TRACKING 344 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 345 select VIRT_CPU_ACCOUNTING 346 select CONTEXT_TRACKING 347 help 348 Select this option to enable task and CPU time accounting on full 349 dynticks systems. This accounting is implemented by watching every 350 kernel-user boundaries using the context tracking subsystem. 351 The accounting is thus performed at the expense of some significant 352 overhead. 353 354 For now this is only useful if you are working on the full 355 dynticks subsystem development. 356 357 If unsure, say N. 358 359config IRQ_TIME_ACCOUNTING 360 bool "Fine granularity task level IRQ time accounting" 361 depends on HAVE_IRQ_TIME_ACCOUNTING && !NO_HZ_FULL 362 help 363 Select this option to enable fine granularity task irq time 364 accounting. This is done by reading a timestamp on each 365 transitions between softirq and hardirq state, so there can be a 366 small performance impact. 367 368 If in doubt, say N here. 369 370endchoice 371 372config BSD_PROCESS_ACCT 373 bool "BSD Process Accounting" 374 help 375 If you say Y here, a user level program will be able to instruct the 376 kernel (via a special system call) to write process accounting 377 information to a file: whenever a process exits, information about 378 that process will be appended to the file by the kernel. The 379 information includes things such as creation time, owning user, 380 command name, memory usage, controlling terminal etc. (the complete 381 list is in the struct acct in <file:include/linux/acct.h>). It is 382 up to the user level program to do useful things with this 383 information. This is generally a good idea, so say Y. 384 385config BSD_PROCESS_ACCT_V3 386 bool "BSD Process Accounting version 3 file format" 387 depends on BSD_PROCESS_ACCT 388 default n 389 help 390 If you say Y here, the process accounting information is written 391 in a new file format that also logs the process IDs of each 392 process and it's parent. Note that this file format is incompatible 393 with previous v0/v1/v2 file formats, so you will need updated tools 394 for processing it. A preliminary version of these tools is available 395 at <http://www.gnu.org/software/acct/>. 396 397config TASKSTATS 398 bool "Export task/process statistics through netlink" 399 depends on NET 400 default n 401 help 402 Export selected statistics for tasks/processes through the 403 generic netlink interface. Unlike BSD process accounting, the 404 statistics are available during the lifetime of tasks/processes as 405 responses to commands. Like BSD accounting, they are sent to user 406 space on task exit. 407 408 Say N if unsure. 409 410config TASK_DELAY_ACCT 411 bool "Enable per-task delay accounting" 412 depends on TASKSTATS 413 help 414 Collect information on time spent by a task waiting for system 415 resources like cpu, synchronous block I/O completion and swapping 416 in pages. Such statistics can help in setting a task's priorities 417 relative to other tasks for cpu, io, rss limits etc. 418 419 Say N if unsure. 420 421config TASK_XACCT 422 bool "Enable extended accounting over taskstats" 423 depends on TASKSTATS 424 help 425 Collect extended task accounting data and send the data 426 to userland for processing over the taskstats interface. 427 428 Say N if unsure. 429 430config TASK_IO_ACCOUNTING 431 bool "Enable per-task storage I/O accounting" 432 depends on TASK_XACCT 433 help 434 Collect information on the number of bytes of storage I/O which this 435 task has caused. 436 437 Say N if unsure. 438 439endmenu # "CPU/Task time and stats accounting" 440 441menu "RCU Subsystem" 442 443choice 444 prompt "RCU Implementation" 445 default TREE_RCU 446 447config TREE_RCU 448 bool "Tree-based hierarchical RCU" 449 depends on !PREEMPT && SMP 450 select IRQ_WORK 451 help 452 This option selects the RCU implementation that is 453 designed for very large SMP system with hundreds or 454 thousands of CPUs. It also scales down nicely to 455 smaller systems. 456 457config TREE_PREEMPT_RCU 458 bool "Preemptible tree-based hierarchical RCU" 459 depends on PREEMPT 460 select IRQ_WORK 461 help 462 This option selects the RCU implementation that is 463 designed for very large SMP systems with hundreds or 464 thousands of CPUs, but for which real-time response 465 is also required. It also scales down nicely to 466 smaller systems. 467 468 Select this option if you are unsure. 469 470config TINY_RCU 471 bool "UP-only small-memory-footprint RCU" 472 depends on !PREEMPT && !SMP 473 help 474 This option selects the RCU implementation that is 475 designed for UP systems from which real-time response 476 is not required. This option greatly reduces the 477 memory footprint of RCU. 478 479endchoice 480 481config PREEMPT_RCU 482 def_bool TREE_PREEMPT_RCU 483 help 484 This option enables preemptible-RCU code that is common between 485 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations. 486 487config RCU_STALL_COMMON 488 def_bool ( TREE_RCU || TREE_PREEMPT_RCU || RCU_TRACE ) 489 help 490 This option enables RCU CPU stall code that is common between 491 the TINY and TREE variants of RCU. The purpose is to allow 492 the tiny variants to disable RCU CPU stall warnings, while 493 making these warnings mandatory for the tree variants. 494 495config CONTEXT_TRACKING 496 bool 497 498config RCU_USER_QS 499 bool "Consider userspace as in RCU extended quiescent state" 500 depends on HAVE_CONTEXT_TRACKING && SMP 501 select CONTEXT_TRACKING 502 help 503 This option sets hooks on kernel / userspace boundaries and 504 puts RCU in extended quiescent state when the CPU runs in 505 userspace. It means that when a CPU runs in userspace, it is 506 excluded from the global RCU state machine and thus doesn't 507 try to keep the timer tick on for RCU. 508 509 Unless you want to hack and help the development of the full 510 dynticks mode, you shouldn't enable this option. It also 511 adds unnecessary overhead. 512 513 If unsure say N 514 515config CONTEXT_TRACKING_FORCE 516 bool "Force context tracking" 517 depends on CONTEXT_TRACKING 518 default y if !NO_HZ_FULL 519 help 520 The major pre-requirement for full dynticks to work is to 521 support the context tracking subsystem. But there are also 522 other dependencies to provide in order to make the full 523 dynticks working. 524 525 This option stands for testing when an arch implements the 526 context tracking backend but doesn't yet fullfill all the 527 requirements to make the full dynticks feature working. 528 Without the full dynticks, there is no way to test the support 529 for context tracking and the subsystems that rely on it: RCU 530 userspace extended quiescent state and tickless cputime 531 accounting. This option copes with the absence of the full 532 dynticks subsystem by forcing the context tracking on all 533 CPUs in the system. 534 535 Say Y only if you're working on the developpement of an 536 architecture backend for the context tracking. 537 538 Say N otherwise, this option brings an overhead that you 539 don't want in production. 540 541 542config RCU_FANOUT 543 int "Tree-based hierarchical RCU fanout value" 544 range 2 64 if 64BIT 545 range 2 32 if !64BIT 546 depends on TREE_RCU || TREE_PREEMPT_RCU 547 default 64 if 64BIT 548 default 32 if !64BIT 549 help 550 This option controls the fanout of hierarchical implementations 551 of RCU, allowing RCU to work efficiently on machines with 552 large numbers of CPUs. This value must be at least the fourth 553 root of NR_CPUS, which allows NR_CPUS to be insanely large. 554 The default value of RCU_FANOUT should be used for production 555 systems, but if you are stress-testing the RCU implementation 556 itself, small RCU_FANOUT values allow you to test large-system 557 code paths on small(er) systems. 558 559 Select a specific number if testing RCU itself. 560 Take the default if unsure. 561 562config RCU_FANOUT_LEAF 563 int "Tree-based hierarchical RCU leaf-level fanout value" 564 range 2 RCU_FANOUT if 64BIT 565 range 2 RCU_FANOUT if !64BIT 566 depends on TREE_RCU || TREE_PREEMPT_RCU 567 default 16 568 help 569 This option controls the leaf-level fanout of hierarchical 570 implementations of RCU, and allows trading off cache misses 571 against lock contention. Systems that synchronize their 572 scheduling-clock interrupts for energy-efficiency reasons will 573 want the default because the smaller leaf-level fanout keeps 574 lock contention levels acceptably low. Very large systems 575 (hundreds or thousands of CPUs) will instead want to set this 576 value to the maximum value possible in order to reduce the 577 number of cache misses incurred during RCU's grace-period 578 initialization. These systems tend to run CPU-bound, and thus 579 are not helped by synchronized interrupts, and thus tend to 580 skew them, which reduces lock contention enough that large 581 leaf-level fanouts work well. 582 583 Select a specific number if testing RCU itself. 584 585 Select the maximum permissible value for large systems. 586 587 Take the default if unsure. 588 589config RCU_FANOUT_EXACT 590 bool "Disable tree-based hierarchical RCU auto-balancing" 591 depends on TREE_RCU || TREE_PREEMPT_RCU 592 default n 593 help 594 This option forces use of the exact RCU_FANOUT value specified, 595 regardless of imbalances in the hierarchy. This is useful for 596 testing RCU itself, and might one day be useful on systems with 597 strong NUMA behavior. 598 599 Without RCU_FANOUT_EXACT, the code will balance the hierarchy. 600 601 Say N if unsure. 602 603config RCU_FAST_NO_HZ 604 bool "Accelerate last non-dyntick-idle CPU's grace periods" 605 depends on NO_HZ_COMMON && SMP 606 default n 607 help 608 This option permits CPUs to enter dynticks-idle state even if 609 they have RCU callbacks queued, and prevents RCU from waking 610 these CPUs up more than roughly once every four jiffies (by 611 default, you can adjust this using the rcutree.rcu_idle_gp_delay 612 parameter), thus improving energy efficiency. On the other 613 hand, this option increases the duration of RCU grace periods, 614 for example, slowing down synchronize_rcu(). 615 616 Say Y if energy efficiency is critically important, and you 617 don't care about increased grace-period durations. 618 619 Say N if you are unsure. 620 621config TREE_RCU_TRACE 622 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU ) 623 select DEBUG_FS 624 help 625 This option provides tracing for the TREE_RCU and 626 TREE_PREEMPT_RCU implementations, permitting Makefile to 627 trivially select kernel/rcutree_trace.c. 628 629config RCU_BOOST 630 bool "Enable RCU priority boosting" 631 depends on RT_MUTEXES && PREEMPT_RCU 632 default n 633 help 634 This option boosts the priority of preempted RCU readers that 635 block the current preemptible RCU grace period for too long. 636 This option also prevents heavy loads from blocking RCU 637 callback invocation for all flavors of RCU. 638 639 Say Y here if you are working with real-time apps or heavy loads 640 Say N here if you are unsure. 641 642config RCU_BOOST_PRIO 643 int "Real-time priority to boost RCU readers to" 644 range 1 99 645 depends on RCU_BOOST 646 default 1 647 help 648 This option specifies the real-time priority to which long-term 649 preempted RCU readers are to be boosted. If you are working 650 with a real-time application that has one or more CPU-bound 651 threads running at a real-time priority level, you should set 652 RCU_BOOST_PRIO to a priority higher then the highest-priority 653 real-time CPU-bound thread. The default RCU_BOOST_PRIO value 654 of 1 is appropriate in the common case, which is real-time 655 applications that do not have any CPU-bound threads. 656 657 Some real-time applications might not have a single real-time 658 thread that saturates a given CPU, but instead might have 659 multiple real-time threads that, taken together, fully utilize 660 that CPU. In this case, you should set RCU_BOOST_PRIO to 661 a priority higher than the lowest-priority thread that is 662 conspiring to prevent the CPU from running any non-real-time 663 tasks. For example, if one thread at priority 10 and another 664 thread at priority 5 are between themselves fully consuming 665 the CPU time on a given CPU, then RCU_BOOST_PRIO should be 666 set to priority 6 or higher. 667 668 Specify the real-time priority, or take the default if unsure. 669 670config RCU_BOOST_DELAY 671 int "Milliseconds to delay boosting after RCU grace-period start" 672 range 0 3000 673 depends on RCU_BOOST 674 default 500 675 help 676 This option specifies the time to wait after the beginning of 677 a given grace period before priority-boosting preempted RCU 678 readers blocking that grace period. Note that any RCU reader 679 blocking an expedited RCU grace period is boosted immediately. 680 681 Accept the default if unsure. 682 683config RCU_NOCB_CPU 684 bool "Offload RCU callback processing from boot-selected CPUs" 685 depends on TREE_RCU || TREE_PREEMPT_RCU 686 default n 687 help 688 Use this option to reduce OS jitter for aggressive HPC or 689 real-time workloads. It can also be used to offload RCU 690 callback invocation to energy-efficient CPUs in battery-powered 691 asymmetric multiprocessors. 692 693 This option offloads callback invocation from the set of 694 CPUs specified at boot time by the rcu_nocbs parameter. 695 For each such CPU, a kthread ("rcuox/N") will be created to 696 invoke callbacks, where the "N" is the CPU being offloaded, 697 and where the "x" is "b" for RCU-bh, "p" for RCU-preempt, and 698 "s" for RCU-sched. Nothing prevents this kthread from running 699 on the specified CPUs, but (1) the kthreads may be preempted 700 between each callback, and (2) affinity or cgroups can be used 701 to force the kthreads to run on whatever set of CPUs is desired. 702 703 Say Y here if you want to help to debug reduced OS jitter. 704 Say N here if you are unsure. 705 706choice 707 prompt "Build-forced no-CBs CPUs" 708 default RCU_NOCB_CPU_NONE 709 help 710 This option allows no-CBs CPUs (whose RCU callbacks are invoked 711 from kthreads rather than from softirq context) to be specified 712 at build time. Additional no-CBs CPUs may be specified by 713 the rcu_nocbs= boot parameter. 714 715config RCU_NOCB_CPU_NONE 716 bool "No build_forced no-CBs CPUs" 717 depends on RCU_NOCB_CPU && !NO_HZ_FULL 718 help 719 This option does not force any of the CPUs to be no-CBs CPUs. 720 Only CPUs designated by the rcu_nocbs= boot parameter will be 721 no-CBs CPUs, whose RCU callbacks will be invoked by per-CPU 722 kthreads whose names begin with "rcuo". All other CPUs will 723 invoke their own RCU callbacks in softirq context. 724 725 Select this option if you want to choose no-CBs CPUs at 726 boot time, for example, to allow testing of different no-CBs 727 configurations without having to rebuild the kernel each time. 728 729config RCU_NOCB_CPU_ZERO 730 bool "CPU 0 is a build_forced no-CBs CPU" 731 depends on RCU_NOCB_CPU && !NO_HZ_FULL 732 help 733 This option forces CPU 0 to be a no-CBs CPU, so that its RCU 734 callbacks are invoked by a per-CPU kthread whose name begins 735 with "rcuo". Additional CPUs may be designated as no-CBs 736 CPUs using the rcu_nocbs= boot parameter will be no-CBs CPUs. 737 All other CPUs will invoke their own RCU callbacks in softirq 738 context. 739 740 Select this if CPU 0 needs to be a no-CBs CPU for real-time 741 or energy-efficiency reasons, but the real reason it exists 742 is to ensure that randconfig testing covers mixed systems. 743 744config RCU_NOCB_CPU_ALL 745 bool "All CPUs are build_forced no-CBs CPUs" 746 depends on RCU_NOCB_CPU 747 help 748 This option forces all CPUs to be no-CBs CPUs. The rcu_nocbs= 749 boot parameter will be ignored. All CPUs' RCU callbacks will 750 be executed in the context of per-CPU rcuo kthreads created for 751 this purpose. Assuming that the kthreads whose names start with 752 "rcuo" are bound to "housekeeping" CPUs, this reduces OS jitter 753 on the remaining CPUs, but might decrease memory locality during 754 RCU-callback invocation, thus potentially degrading throughput. 755 756 Select this if all CPUs need to be no-CBs CPUs for real-time 757 or energy-efficiency reasons. 758 759endchoice 760 761endmenu # "RCU Subsystem" 762 763config IKCONFIG 764 tristate "Kernel .config support" 765 ---help--- 766 This option enables the complete Linux kernel ".config" file 767 contents to be saved in the kernel. It provides documentation 768 of which kernel options are used in a running kernel or in an 769 on-disk kernel. This information can be extracted from the kernel 770 image file with the script scripts/extract-ikconfig and used as 771 input to rebuild the current kernel or to build another kernel. 772 It can also be extracted from a running kernel by reading 773 /proc/config.gz if enabled (below). 774 775config IKCONFIG_PROC 776 bool "Enable access to .config through /proc/config.gz" 777 depends on IKCONFIG && PROC_FS 778 ---help--- 779 This option enables access to the kernel configuration file 780 through /proc/config.gz. 781 782config LOG_BUF_SHIFT 783 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 784 range 12 21 785 default 17 786 help 787 Select kernel log buffer size as a power of 2. 788 Examples: 789 17 => 128 KB 790 16 => 64 KB 791 15 => 32 KB 792 14 => 16 KB 793 13 => 8 KB 794 12 => 4 KB 795 796# 797# Architectures with an unreliable sched_clock() should select this: 798# 799config HAVE_UNSTABLE_SCHED_CLOCK 800 bool 801 802config GENERIC_SCHED_CLOCK 803 bool 804 805# 806# For architectures that want to enable the support for NUMA-affine scheduler 807# balancing logic: 808# 809config ARCH_SUPPORTS_NUMA_BALANCING 810 bool 811 812# 813# For architectures that know their GCC __int128 support is sound 814# 815config ARCH_SUPPORTS_INT128 816 bool 817 818# For architectures that (ab)use NUMA to represent different memory regions 819# all cpu-local but of different latencies, such as SuperH. 820# 821config ARCH_WANT_NUMA_VARIABLE_LOCALITY 822 bool 823 824# 825# For architectures that are willing to define _PAGE_NUMA as _PAGE_PROTNONE 826config ARCH_WANTS_PROT_NUMA_PROT_NONE 827 bool 828 829config ARCH_USES_NUMA_PROT_NONE 830 bool 831 default y 832 depends on ARCH_WANTS_PROT_NUMA_PROT_NONE 833 depends on NUMA_BALANCING 834 835config NUMA_BALANCING_DEFAULT_ENABLED 836 bool "Automatically enable NUMA aware memory/task placement" 837 default y 838 depends on NUMA_BALANCING 839 help 840 If set, automatic NUMA balancing will be enabled if running on a NUMA 841 machine. 842 843config NUMA_BALANCING 844 bool "Memory placement aware NUMA scheduler" 845 depends on ARCH_SUPPORTS_NUMA_BALANCING 846 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 847 depends on SMP && NUMA && MIGRATION 848 help 849 This option adds support for automatic NUMA aware memory/task placement. 850 The mechanism is quite primitive and is based on migrating memory when 851 it has references to the node the task is running on. 852 853 This system will be inactive on UMA systems. 854 855menuconfig CGROUPS 856 boolean "Control Group support" 857 depends on EVENTFD 858 help 859 This option adds support for grouping sets of processes together, for 860 use with process control subsystems such as Cpusets, CFS, memory 861 controls or device isolation. 862 See 863 - Documentation/scheduler/sched-design-CFS.txt (CFS) 864 - Documentation/cgroups/ (features for grouping, isolation 865 and resource control) 866 867 Say N if unsure. 868 869if CGROUPS 870 871config CGROUP_DEBUG 872 bool "Example debug cgroup subsystem" 873 default n 874 help 875 This option enables a simple cgroup subsystem that 876 exports useful debugging information about the cgroups 877 framework. 878 879 Say N if unsure. 880 881config CGROUP_FREEZER 882 bool "Freezer cgroup subsystem" 883 help 884 Provides a way to freeze and unfreeze all tasks in a 885 cgroup. 886 887config CGROUP_DEVICE 888 bool "Device controller for cgroups" 889 help 890 Provides a cgroup implementing whitelists for devices which 891 a process in the cgroup can mknod or open. 892 893config CPUSETS 894 bool "Cpuset support" 895 help 896 This option will let you create and manage CPUSETs which 897 allow dynamically partitioning a system into sets of CPUs and 898 Memory Nodes and assigning tasks to run only within those sets. 899 This is primarily useful on large SMP or NUMA systems. 900 901 Say N if unsure. 902 903config PROC_PID_CPUSET 904 bool "Include legacy /proc/<pid>/cpuset file" 905 depends on CPUSETS 906 default y 907 908config CGROUP_CPUACCT 909 bool "Simple CPU accounting cgroup subsystem" 910 help 911 Provides a simple Resource Controller for monitoring the 912 total CPU consumed by the tasks in a cgroup. 913 914config RESOURCE_COUNTERS 915 bool "Resource counters" 916 help 917 This option enables controller independent resource accounting 918 infrastructure that works with cgroups. 919 920config MEMCG 921 bool "Memory Resource Controller for Control Groups" 922 depends on RESOURCE_COUNTERS 923 select MM_OWNER 924 help 925 Provides a memory resource controller that manages both anonymous 926 memory and page cache. (See Documentation/cgroups/memory.txt) 927 928 Note that setting this option increases fixed memory overhead 929 associated with each page of memory in the system. By this, 930 8(16)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory 931 usage tracking struct at boot. Total amount of this is printed out 932 at boot. 933 934 Only enable when you're ok with these trade offs and really 935 sure you need the memory resource controller. Even when you enable 936 this, you can set "cgroup_disable=memory" at your boot option to 937 disable memory resource controller and you can avoid overheads. 938 (and lose benefits of memory resource controller) 939 940 This config option also selects MM_OWNER config option, which 941 could in turn add some fork/exit overhead. 942 943config MEMCG_SWAP 944 bool "Memory Resource Controller Swap Extension" 945 depends on MEMCG && SWAP 946 help 947 Add swap management feature to memory resource controller. When you 948 enable this, you can limit mem+swap usage per cgroup. In other words, 949 when you disable this, memory resource controller has no cares to 950 usage of swap...a process can exhaust all of the swap. This extension 951 is useful when you want to avoid exhaustion swap but this itself 952 adds more overheads and consumes memory for remembering information. 953 Especially if you use 32bit system or small memory system, please 954 be careful about enabling this. When memory resource controller 955 is disabled by boot option, this will be automatically disabled and 956 there will be no overhead from this. Even when you set this config=y, 957 if boot option "swapaccount=0" is set, swap will not be accounted. 958 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page 959 size is 4096bytes, 512k per 1Gbytes of swap. 960config MEMCG_SWAP_ENABLED 961 bool "Memory Resource Controller Swap Extension enabled by default" 962 depends on MEMCG_SWAP 963 default y 964 help 965 Memory Resource Controller Swap Extension comes with its price in 966 a bigger memory consumption. General purpose distribution kernels 967 which want to enable the feature but keep it disabled by default 968 and let the user enable it by swapaccount=1 boot command line 969 parameter should have this option unselected. 970 For those who want to have the feature enabled by default should 971 select this option (if, for some reason, they need to disable it 972 then swapaccount=0 does the trick). 973config MEMCG_KMEM 974 bool "Memory Resource Controller Kernel Memory accounting" 975 depends on MEMCG 976 depends on SLUB || SLAB 977 help 978 The Kernel Memory extension for Memory Resource Controller can limit 979 the amount of memory used by kernel objects in the system. Those are 980 fundamentally different from the entities handled by the standard 981 Memory Controller, which are page-based, and can be swapped. Users of 982 the kmem extension can use it to guarantee that no group of processes 983 will ever exhaust kernel resources alone. 984 985config CGROUP_HUGETLB 986 bool "HugeTLB Resource Controller for Control Groups" 987 depends on RESOURCE_COUNTERS && HUGETLB_PAGE 988 default n 989 help 990 Provides a cgroup Resource Controller for HugeTLB pages. 991 When you enable this, you can put a per cgroup limit on HugeTLB usage. 992 The limit is enforced during page fault. Since HugeTLB doesn't 993 support page reclaim, enforcing the limit at page fault time implies 994 that, the application will get SIGBUS signal if it tries to access 995 HugeTLB pages beyond its limit. This requires the application to know 996 beforehand how much HugeTLB pages it would require for its use. The 997 control group is tracked in the third page lru pointer. This means 998 that we cannot use the controller with huge page less than 3 pages. 999 1000config CGROUP_PERF 1001 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring" 1002 depends on PERF_EVENTS && CGROUPS 1003 help 1004 This option extends the per-cpu mode to restrict monitoring to 1005 threads which belong to the cgroup specified and run on the 1006 designated cpu. 1007 1008 Say N if unsure. 1009 1010menuconfig CGROUP_SCHED 1011 bool "Group CPU scheduler" 1012 default n 1013 help 1014 This feature lets CPU scheduler recognize task groups and control CPU 1015 bandwidth allocation to such task groups. It uses cgroups to group 1016 tasks. 1017 1018if CGROUP_SCHED 1019config FAIR_GROUP_SCHED 1020 bool "Group scheduling for SCHED_OTHER" 1021 depends on CGROUP_SCHED 1022 default CGROUP_SCHED 1023 1024config CFS_BANDWIDTH 1025 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 1026 depends on FAIR_GROUP_SCHED 1027 default n 1028 help 1029 This option allows users to define CPU bandwidth rates (limits) for 1030 tasks running within the fair group scheduler. Groups with no limit 1031 set are considered to be unconstrained and will run with no 1032 restriction. 1033 See tip/Documentation/scheduler/sched-bwc.txt for more information. 1034 1035config RT_GROUP_SCHED 1036 bool "Group scheduling for SCHED_RR/FIFO" 1037 depends on CGROUP_SCHED 1038 default n 1039 help 1040 This feature lets you explicitly allocate real CPU bandwidth 1041 to task groups. If enabled, it will also make it impossible to 1042 schedule realtime tasks for non-root users until you allocate 1043 realtime bandwidth for them. 1044 See Documentation/scheduler/sched-rt-group.txt for more information. 1045 1046endif #CGROUP_SCHED 1047 1048config BLK_CGROUP 1049 bool "Block IO controller" 1050 depends on BLOCK 1051 default n 1052 ---help--- 1053 Generic block IO controller cgroup interface. This is the common 1054 cgroup interface which should be used by various IO controlling 1055 policies. 1056 1057 Currently, CFQ IO scheduler uses it to recognize task groups and 1058 control disk bandwidth allocation (proportional time slice allocation) 1059 to such task groups. It is also used by bio throttling logic in 1060 block layer to implement upper limit in IO rates on a device. 1061 1062 This option only enables generic Block IO controller infrastructure. 1063 One needs to also enable actual IO controlling logic/policy. For 1064 enabling proportional weight division of disk bandwidth in CFQ, set 1065 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 1066 CONFIG_BLK_DEV_THROTTLING=y. 1067 1068 See Documentation/cgroups/blkio-controller.txt for more information. 1069 1070config DEBUG_BLK_CGROUP 1071 bool "Enable Block IO controller debugging" 1072 depends on BLK_CGROUP 1073 default n 1074 ---help--- 1075 Enable some debugging help. Currently it exports additional stat 1076 files in a cgroup which can be useful for debugging. 1077 1078endif # CGROUPS 1079 1080config CHECKPOINT_RESTORE 1081 bool "Checkpoint/restore support" if EXPERT 1082 default n 1083 help 1084 Enables additional kernel features in a sake of checkpoint/restore. 1085 In particular it adds auxiliary prctl codes to setup process text, 1086 data and heap segment sizes, and a few additional /proc filesystem 1087 entries. 1088 1089 If unsure, say N here. 1090 1091menuconfig NAMESPACES 1092 bool "Namespaces support" if EXPERT 1093 default !EXPERT 1094 help 1095 Provides the way to make tasks work with different objects using 1096 the same id. For example same IPC id may refer to different objects 1097 or same user id or pid may refer to different tasks when used in 1098 different namespaces. 1099 1100if NAMESPACES 1101 1102config UTS_NS 1103 bool "UTS namespace" 1104 default y 1105 help 1106 In this namespace tasks see different info provided with the 1107 uname() system call 1108 1109config IPC_NS 1110 bool "IPC namespace" 1111 depends on (SYSVIPC || POSIX_MQUEUE) 1112 default y 1113 help 1114 In this namespace tasks work with IPC ids which correspond to 1115 different IPC objects in different namespaces. 1116 1117config USER_NS 1118 bool "User namespace" 1119 select UIDGID_STRICT_TYPE_CHECKS 1120 1121 default n 1122 help 1123 This allows containers, i.e. vservers, to use user namespaces 1124 to provide different user info for different servers. 1125 1126 When user namespaces are enabled in the kernel it is 1127 recommended that the MEMCG and MEMCG_KMEM options also be 1128 enabled and that user-space use the memory control groups to 1129 limit the amount of memory a memory unprivileged users can 1130 use. 1131 1132 If unsure, say N. 1133 1134config PID_NS 1135 bool "PID Namespaces" 1136 default y 1137 help 1138 Support process id namespaces. This allows having multiple 1139 processes with the same pid as long as they are in different 1140 pid namespaces. This is a building block of containers. 1141 1142config NET_NS 1143 bool "Network namespace" 1144 depends on NET 1145 default y 1146 help 1147 Allow user space to create what appear to be multiple instances 1148 of the network stack. 1149 1150endif # NAMESPACES 1151 1152config UIDGID_STRICT_TYPE_CHECKS 1153 bool "Require conversions between uid/gids and their internal representation" 1154 default n 1155 help 1156 While the nececessary conversions are being added to all subsystems this option allows 1157 the code to continue to build for unconverted subsystems. 1158 1159 Say Y here if you want the strict type checking enabled 1160 1161config SCHED_AUTOGROUP 1162 bool "Automatic process group scheduling" 1163 select EVENTFD 1164 select CGROUPS 1165 select CGROUP_SCHED 1166 select FAIR_GROUP_SCHED 1167 help 1168 This option optimizes the scheduler for common desktop workloads by 1169 automatically creating and populating task groups. This separation 1170 of workloads isolates aggressive CPU burners (like build jobs) from 1171 desktop applications. Task group autogeneration is currently based 1172 upon task session. 1173 1174config MM_OWNER 1175 bool 1176 1177config SYSFS_DEPRECATED 1178 bool "Enable deprecated sysfs features to support old userspace tools" 1179 depends on SYSFS 1180 default n 1181 help 1182 This option adds code that switches the layout of the "block" class 1183 devices, to not show up in /sys/class/block/, but only in 1184 /sys/block/. 1185 1186 This switch is only active when the sysfs.deprecated=1 boot option is 1187 passed or the SYSFS_DEPRECATED_V2 option is set. 1188 1189 This option allows new kernels to run on old distributions and tools, 1190 which might get confused by /sys/class/block/. Since 2007/2008 all 1191 major distributions and tools handle this just fine. 1192 1193 Recent distributions and userspace tools after 2009/2010 depend on 1194 the existence of /sys/class/block/, and will not work with this 1195 option enabled. 1196 1197 Only if you are using a new kernel on an old distribution, you might 1198 need to say Y here. 1199 1200config SYSFS_DEPRECATED_V2 1201 bool "Enable deprecated sysfs features by default" 1202 default n 1203 depends on SYSFS 1204 depends on SYSFS_DEPRECATED 1205 help 1206 Enable deprecated sysfs by default. 1207 1208 See the CONFIG_SYSFS_DEPRECATED option for more details about this 1209 option. 1210 1211 Only if you are using a new kernel on an old distribution, you might 1212 need to say Y here. Even then, odds are you would not need it 1213 enabled, you can always pass the boot option if absolutely necessary. 1214 1215config RELAY 1216 bool "Kernel->user space relay support (formerly relayfs)" 1217 help 1218 This option enables support for relay interface support in 1219 certain file systems (such as debugfs). 1220 It is designed to provide an efficient mechanism for tools and 1221 facilities to relay large amounts of data from kernel space to 1222 user space. 1223 1224 If unsure, say N. 1225 1226config BLK_DEV_INITRD 1227 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1228 depends on BROKEN || !FRV 1229 help 1230 The initial RAM filesystem is a ramfs which is loaded by the 1231 boot loader (loadlin or lilo) and that is mounted as root 1232 before the normal boot procedure. It is typically used to 1233 load modules needed to mount the "real" root file system, 1234 etc. See <file:Documentation/initrd.txt> for details. 1235 1236 If RAM disk support (BLK_DEV_RAM) is also included, this 1237 also enables initial RAM disk (initrd) support and adds 1238 15 Kbytes (more on some other architectures) to the kernel size. 1239 1240 If unsure say Y. 1241 1242if BLK_DEV_INITRD 1243 1244source "usr/Kconfig" 1245 1246endif 1247 1248config CC_OPTIMIZE_FOR_SIZE 1249 bool "Optimize for size" 1250 help 1251 Enabling this option will pass "-Os" instead of "-O2" to gcc 1252 resulting in a smaller kernel. 1253 1254 If unsure, say N. 1255 1256config SYSCTL 1257 bool 1258 1259config ANON_INODES 1260 bool 1261 1262config HAVE_UID16 1263 bool 1264 1265config SYSCTL_EXCEPTION_TRACE 1266 bool 1267 help 1268 Enable support for /proc/sys/debug/exception-trace. 1269 1270config SYSCTL_ARCH_UNALIGN_NO_WARN 1271 bool 1272 help 1273 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1274 Allows arch to define/use @no_unaligned_warning to possibly warn 1275 about unaligned access emulation going on under the hood. 1276 1277config SYSCTL_ARCH_UNALIGN_ALLOW 1278 bool 1279 help 1280 Enable support for /proc/sys/kernel/unaligned-trap 1281 Allows arches to define/use @unaligned_enabled to runtime toggle 1282 the unaligned access emulation. 1283 see arch/parisc/kernel/unaligned.c for reference 1284 1285config HAVE_PCSPKR_PLATFORM 1286 bool 1287 1288menuconfig EXPERT 1289 bool "Configure standard kernel features (expert users)" 1290 # Unhide debug options, to make the on-by-default options visible 1291 select DEBUG_KERNEL 1292 help 1293 This option allows certain base kernel options and settings 1294 to be disabled or tweaked. This is for specialized 1295 environments which can tolerate a "non-standard" kernel. 1296 Only use this if you really know what you are doing. 1297 1298config UID16 1299 bool "Enable 16-bit UID system calls" if EXPERT 1300 depends on HAVE_UID16 1301 default y 1302 help 1303 This enables the legacy 16-bit UID syscall wrappers. 1304 1305config SYSCTL_SYSCALL 1306 bool "Sysctl syscall support" if EXPERT 1307 depends on PROC_SYSCTL 1308 default n 1309 select SYSCTL 1310 ---help--- 1311 sys_sysctl uses binary paths that have been found challenging 1312 to properly maintain and use. The interface in /proc/sys 1313 using paths with ascii names is now the primary path to this 1314 information. 1315 1316 Almost nothing using the binary sysctl interface so if you are 1317 trying to save some space it is probably safe to disable this, 1318 making your kernel marginally smaller. 1319 1320 If unsure say N here. 1321 1322config KALLSYMS 1323 bool "Load all symbols for debugging/ksymoops" if EXPERT 1324 default y 1325 help 1326 Say Y here to let the kernel print out symbolic crash information and 1327 symbolic stack backtraces. This increases the size of the kernel 1328 somewhat, as all symbols have to be loaded into the kernel image. 1329 1330config KALLSYMS_ALL 1331 bool "Include all symbols in kallsyms" 1332 depends on DEBUG_KERNEL && KALLSYMS 1333 help 1334 Normally kallsyms only contains the symbols of functions for nicer 1335 OOPS messages and backtraces (i.e., symbols from the text and inittext 1336 sections). This is sufficient for most cases. And only in very rare 1337 cases (e.g., when a debugger is used) all symbols are required (e.g., 1338 names of variables from the data sections, etc). 1339 1340 This option makes sure that all symbols are loaded into the kernel 1341 image (i.e., symbols from all sections) in cost of increased kernel 1342 size (depending on the kernel configuration, it may be 300KiB or 1343 something like this). 1344 1345 Say N unless you really need all symbols. 1346 1347config PRINTK 1348 default y 1349 bool "Enable support for printk" if EXPERT 1350 select IRQ_WORK 1351 help 1352 This option enables normal printk support. Removing it 1353 eliminates most of the message strings from the kernel image 1354 and makes the kernel more or less silent. As this makes it 1355 very difficult to diagnose system problems, saying N here is 1356 strongly discouraged. 1357 1358config BUG 1359 bool "BUG() support" if EXPERT 1360 default y 1361 help 1362 Disabling this option eliminates support for BUG and WARN, reducing 1363 the size of your kernel image and potentially quietly ignoring 1364 numerous fatal conditions. You should only consider disabling this 1365 option for embedded systems with no facilities for reporting errors. 1366 Just say Y. 1367 1368config ELF_CORE 1369 depends on COREDUMP 1370 default y 1371 bool "Enable ELF core dumps" if EXPERT 1372 help 1373 Enable support for generating core dumps. Disabling saves about 4k. 1374 1375 1376config PCSPKR_PLATFORM 1377 bool "Enable PC-Speaker support" if EXPERT 1378 depends on HAVE_PCSPKR_PLATFORM 1379 select I8253_LOCK 1380 default y 1381 help 1382 This option allows to disable the internal PC-Speaker 1383 support, saving some memory. 1384 1385config BASE_FULL 1386 default y 1387 bool "Enable full-sized data structures for core" if EXPERT 1388 help 1389 Disabling this option reduces the size of miscellaneous core 1390 kernel data structures. This saves memory on small machines, 1391 but may reduce performance. 1392 1393config FUTEX 1394 bool "Enable futex support" if EXPERT 1395 default y 1396 select RT_MUTEXES 1397 help 1398 Disabling this option will cause the kernel to be built without 1399 support for "fast userspace mutexes". The resulting kernel may not 1400 run glibc-based applications correctly. 1401 1402config EPOLL 1403 bool "Enable eventpoll support" if EXPERT 1404 default y 1405 select ANON_INODES 1406 help 1407 Disabling this option will cause the kernel to be built without 1408 support for epoll family of system calls. 1409 1410config SIGNALFD 1411 bool "Enable signalfd() system call" if EXPERT 1412 select ANON_INODES 1413 default y 1414 help 1415 Enable the signalfd() system call that allows to receive signals 1416 on a file descriptor. 1417 1418 If unsure, say Y. 1419 1420config TIMERFD 1421 bool "Enable timerfd() system call" if EXPERT 1422 select ANON_INODES 1423 default y 1424 help 1425 Enable the timerfd() system call that allows to receive timer 1426 events on a file descriptor. 1427 1428 If unsure, say Y. 1429 1430config EVENTFD 1431 bool "Enable eventfd() system call" if EXPERT 1432 select ANON_INODES 1433 default y 1434 help 1435 Enable the eventfd() system call that allows to receive both 1436 kernel notification (ie. KAIO) or userspace notifications. 1437 1438 If unsure, say Y. 1439 1440config SHMEM 1441 bool "Use full shmem filesystem" if EXPERT 1442 default y 1443 depends on MMU 1444 help 1445 The shmem is an internal filesystem used to manage shared memory. 1446 It is backed by swap and manages resource limits. It is also exported 1447 to userspace as tmpfs if TMPFS is enabled. Disabling this 1448 option replaces shmem and tmpfs with the much simpler ramfs code, 1449 which may be appropriate on small systems without swap. 1450 1451config AIO 1452 bool "Enable AIO support" if EXPERT 1453 default y 1454 help 1455 This option enables POSIX asynchronous I/O which may by used 1456 by some high performance threaded applications. Disabling 1457 this option saves about 7k. 1458 1459config PCI_QUIRKS 1460 default y 1461 bool "Enable PCI quirk workarounds" if EXPERT 1462 depends on PCI 1463 help 1464 This enables workarounds for various PCI chipset 1465 bugs/quirks. Disable this only if your target machine is 1466 unaffected by PCI quirks. 1467 1468config EMBEDDED 1469 bool "Embedded system" 1470 select EXPERT 1471 help 1472 This option should be enabled if compiling the kernel for 1473 an embedded system so certain expert options are available 1474 for configuration. 1475 1476config HAVE_PERF_EVENTS 1477 bool 1478 help 1479 See tools/perf/design.txt for details. 1480 1481config PERF_USE_VMALLOC 1482 bool 1483 help 1484 See tools/perf/design.txt for details 1485 1486menu "Kernel Performance Events And Counters" 1487 1488config PERF_EVENTS 1489 bool "Kernel performance events and counters" 1490 default y if PROFILING 1491 depends on HAVE_PERF_EVENTS 1492 select ANON_INODES 1493 select IRQ_WORK 1494 help 1495 Enable kernel support for various performance events provided 1496 by software and hardware. 1497 1498 Software events are supported either built-in or via the 1499 use of generic tracepoints. 1500 1501 Most modern CPUs support performance events via performance 1502 counter registers. These registers count the number of certain 1503 types of hw events: such as instructions executed, cachemisses 1504 suffered, or branches mis-predicted - without slowing down the 1505 kernel or applications. These registers can also trigger interrupts 1506 when a threshold number of events have passed - and can thus be 1507 used to profile the code that runs on that CPU. 1508 1509 The Linux Performance Event subsystem provides an abstraction of 1510 these software and hardware event capabilities, available via a 1511 system call and used by the "perf" utility in tools/perf/. It 1512 provides per task and per CPU counters, and it provides event 1513 capabilities on top of those. 1514 1515 Say Y if unsure. 1516 1517config DEBUG_PERF_USE_VMALLOC 1518 default n 1519 bool "Debug: use vmalloc to back perf mmap() buffers" 1520 depends on PERF_EVENTS && DEBUG_KERNEL 1521 select PERF_USE_VMALLOC 1522 help 1523 Use vmalloc memory to back perf mmap() buffers. 1524 1525 Mostly useful for debugging the vmalloc code on platforms 1526 that don't require it. 1527 1528 Say N if unsure. 1529 1530endmenu 1531 1532config VM_EVENT_COUNTERS 1533 default y 1534 bool "Enable VM event counters for /proc/vmstat" if EXPERT 1535 help 1536 VM event counters are needed for event counts to be shown. 1537 This option allows the disabling of the VM event counters 1538 on EXPERT systems. /proc/vmstat will only show page counts 1539 if VM event counters are disabled. 1540 1541config SLUB_DEBUG 1542 default y 1543 bool "Enable SLUB debugging support" if EXPERT 1544 depends on SLUB && SYSFS 1545 help 1546 SLUB has extensive debug support features. Disabling these can 1547 result in significant savings in code size. This also disables 1548 SLUB sysfs support. /sys/slab will not exist and there will be 1549 no support for cache validation etc. 1550 1551config COMPAT_BRK 1552 bool "Disable heap randomization" 1553 default y 1554 help 1555 Randomizing heap placement makes heap exploits harder, but it 1556 also breaks ancient binaries (including anything libc5 based). 1557 This option changes the bootup default to heap randomization 1558 disabled, and can be overridden at runtime by setting 1559 /proc/sys/kernel/randomize_va_space to 2. 1560 1561 On non-ancient distros (post-2000 ones) N is usually a safe choice. 1562 1563choice 1564 prompt "Choose SLAB allocator" 1565 default SLUB 1566 help 1567 This option allows to select a slab allocator. 1568 1569config SLAB 1570 bool "SLAB" 1571 help 1572 The regular slab allocator that is established and known to work 1573 well in all environments. It organizes cache hot objects in 1574 per cpu and per node queues. 1575 1576config SLUB 1577 bool "SLUB (Unqueued Allocator)" 1578 help 1579 SLUB is a slab allocator that minimizes cache line usage 1580 instead of managing queues of cached objects (SLAB approach). 1581 Per cpu caching is realized using slabs of objects instead 1582 of queues of objects. SLUB can use memory efficiently 1583 and has enhanced diagnostics. SLUB is the default choice for 1584 a slab allocator. 1585 1586config SLOB 1587 depends on EXPERT 1588 bool "SLOB (Simple Allocator)" 1589 help 1590 SLOB replaces the stock allocator with a drastically simpler 1591 allocator. SLOB is generally more space efficient but 1592 does not perform as well on large systems. 1593 1594endchoice 1595 1596config SLUB_CPU_PARTIAL 1597 default y 1598 depends on SLUB && SMP 1599 bool "SLUB per cpu partial cache" 1600 help 1601 Per cpu partial caches accellerate objects allocation and freeing 1602 that is local to a processor at the price of more indeterminism 1603 in the latency of the free. On overflow these caches will be cleared 1604 which requires the taking of locks that may cause latency spikes. 1605 Typically one would choose no for a realtime system. 1606 1607config MMAP_ALLOW_UNINITIALIZED 1608 bool "Allow mmapped anonymous memory to be uninitialized" 1609 depends on EXPERT && !MMU 1610 default n 1611 help 1612 Normally, and according to the Linux spec, anonymous memory obtained 1613 from mmap() has it's contents cleared before it is passed to 1614 userspace. Enabling this config option allows you to request that 1615 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1616 providing a huge performance boost. If this option is not enabled, 1617 then the flag will be ignored. 1618 1619 This is taken advantage of by uClibc's malloc(), and also by 1620 ELF-FDPIC binfmt's brk and stack allocator. 1621 1622 Because of the obvious security issues, this option should only be 1623 enabled on embedded devices where you control what is run in 1624 userspace. Since that isn't generally a problem on no-MMU systems, 1625 it is normally safe to say Y here. 1626 1627 See Documentation/nommu-mmap.txt for more information. 1628 1629config PROFILING 1630 bool "Profiling support" 1631 help 1632 Say Y here to enable the extended profiling support mechanisms used 1633 by profilers such as OProfile. 1634 1635# 1636# Place an empty function call at each tracepoint site. Can be 1637# dynamically changed for a probe function. 1638# 1639config TRACEPOINTS 1640 bool 1641 1642source "arch/Kconfig" 1643 1644endmenu # General setup 1645 1646config HAVE_GENERIC_DMA_COHERENT 1647 bool 1648 default n 1649 1650config SLABINFO 1651 bool 1652 depends on PROC_FS 1653 depends on SLAB || SLUB_DEBUG 1654 default y 1655 1656config RT_MUTEXES 1657 boolean 1658 1659config BASE_SMALL 1660 int 1661 default 0 if BASE_FULL 1662 default 1 if !BASE_FULL 1663 1664config SYSTEM_TRUSTED_KEYRING 1665 bool "Provide system-wide ring of trusted keys" 1666 depends on KEYS 1667 help 1668 Provide a system keyring to which trusted keys can be added. Keys in 1669 the keyring are considered to be trusted. Keys may be added at will 1670 by the kernel from compiled-in data and from hardware key stores, but 1671 userspace may only add extra keys if those keys can be verified by 1672 keys already in the keyring. 1673 1674 Keys in this keyring are used by module signature checking. 1675 1676menuconfig MODULES 1677 bool "Enable loadable module support" 1678 option modules 1679 help 1680 Kernel modules are small pieces of compiled code which can 1681 be inserted in the running kernel, rather than being 1682 permanently built into the kernel. You use the "modprobe" 1683 tool to add (and sometimes remove) them. If you say Y here, 1684 many parts of the kernel can be built as modules (by 1685 answering M instead of Y where indicated): this is most 1686 useful for infrequently used options which are not required 1687 for booting. For more information, see the man pages for 1688 modprobe, lsmod, modinfo, insmod and rmmod. 1689 1690 If you say Y here, you will need to run "make 1691 modules_install" to put the modules under /lib/modules/ 1692 where modprobe can find them (you may need to be root to do 1693 this). 1694 1695 If unsure, say Y. 1696 1697if MODULES 1698 1699config MODULE_FORCE_LOAD 1700 bool "Forced module loading" 1701 default n 1702 help 1703 Allow loading of modules without version information (ie. modprobe 1704 --force). Forced module loading sets the 'F' (forced) taint flag and 1705 is usually a really bad idea. 1706 1707config MODULE_UNLOAD 1708 bool "Module unloading" 1709 help 1710 Without this option you will not be able to unload any 1711 modules (note that some modules may not be unloadable 1712 anyway), which makes your kernel smaller, faster 1713 and simpler. If unsure, say Y. 1714 1715config MODULE_FORCE_UNLOAD 1716 bool "Forced module unloading" 1717 depends on MODULE_UNLOAD 1718 help 1719 This option allows you to force a module to unload, even if the 1720 kernel believes it is unsafe: the kernel will remove the module 1721 without waiting for anyone to stop using it (using the -f option to 1722 rmmod). This is mainly for kernel developers and desperate users. 1723 If unsure, say N. 1724 1725config MODVERSIONS 1726 bool "Module versioning support" 1727 help 1728 Usually, you have to use modules compiled with your kernel. 1729 Saying Y here makes it sometimes possible to use modules 1730 compiled for different kernels, by adding enough information 1731 to the modules to (hopefully) spot any changes which would 1732 make them incompatible with the kernel you are running. If 1733 unsure, say N. 1734 1735config MODULE_SRCVERSION_ALL 1736 bool "Source checksum for all modules" 1737 help 1738 Modules which contain a MODULE_VERSION get an extra "srcversion" 1739 field inserted into their modinfo section, which contains a 1740 sum of the source files which made it. This helps maintainers 1741 see exactly which source was used to build a module (since 1742 others sometimes change the module source without updating 1743 the version). With this option, such a "srcversion" field 1744 will be created for all modules. If unsure, say N. 1745 1746config MODULE_SIG 1747 bool "Module signature verification" 1748 depends on MODULES 1749 select SYSTEM_TRUSTED_KEYRING 1750 select KEYS 1751 select CRYPTO 1752 select ASYMMETRIC_KEY_TYPE 1753 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1754 select PUBLIC_KEY_ALGO_RSA 1755 select ASN1 1756 select OID_REGISTRY 1757 select X509_CERTIFICATE_PARSER 1758 help 1759 Check modules for valid signatures upon load: the signature 1760 is simply appended to the module. For more information see 1761 Documentation/module-signing.txt. 1762 1763 !!!WARNING!!! If you enable this option, you MUST make sure that the 1764 module DOES NOT get stripped after being signed. This includes the 1765 debuginfo strip done by some packagers (such as rpmbuild) and 1766 inclusion into an initramfs that wants the module size reduced. 1767 1768config MODULE_SIG_FORCE 1769 bool "Require modules to be validly signed" 1770 depends on MODULE_SIG 1771 help 1772 Reject unsigned modules or signed modules for which we don't have a 1773 key. Without this, such modules will simply taint the kernel. 1774 1775config MODULE_SIG_ALL 1776 bool "Automatically sign all modules" 1777 default y 1778 depends on MODULE_SIG 1779 help 1780 Sign all modules during make modules_install. Without this option, 1781 modules must be signed manually, using the scripts/sign-file tool. 1782 1783comment "Do not forget to sign required modules with scripts/sign-file" 1784 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 1785 1786choice 1787 prompt "Which hash algorithm should modules be signed with?" 1788 depends on MODULE_SIG 1789 help 1790 This determines which sort of hashing algorithm will be used during 1791 signature generation. This algorithm _must_ be built into the kernel 1792 directly so that signature verification can take place. It is not 1793 possible to load a signed module containing the algorithm to check 1794 the signature on that module. 1795 1796config MODULE_SIG_SHA1 1797 bool "Sign modules with SHA-1" 1798 select CRYPTO_SHA1 1799 1800config MODULE_SIG_SHA224 1801 bool "Sign modules with SHA-224" 1802 select CRYPTO_SHA256 1803 1804config MODULE_SIG_SHA256 1805 bool "Sign modules with SHA-256" 1806 select CRYPTO_SHA256 1807 1808config MODULE_SIG_SHA384 1809 bool "Sign modules with SHA-384" 1810 select CRYPTO_SHA512 1811 1812config MODULE_SIG_SHA512 1813 bool "Sign modules with SHA-512" 1814 select CRYPTO_SHA512 1815 1816endchoice 1817 1818config MODULE_SIG_HASH 1819 string 1820 depends on MODULE_SIG 1821 default "sha1" if MODULE_SIG_SHA1 1822 default "sha224" if MODULE_SIG_SHA224 1823 default "sha256" if MODULE_SIG_SHA256 1824 default "sha384" if MODULE_SIG_SHA384 1825 default "sha512" if MODULE_SIG_SHA512 1826 1827endif # MODULES 1828 1829config INIT_ALL_POSSIBLE 1830 bool 1831 help 1832 Back when each arch used to define their own cpu_online_mask and 1833 cpu_possible_mask, some of them chose to initialize cpu_possible_mask 1834 with all 1s, and others with all 0s. When they were centralised, 1835 it was better to provide this option than to break all the archs 1836 and have several arch maintainers pursuing me down dark alleys. 1837 1838config STOP_MACHINE 1839 bool 1840 default y 1841 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU 1842 help 1843 Need stop_machine() primitive. 1844 1845source "block/Kconfig" 1846 1847config PREEMPT_NOTIFIERS 1848 bool 1849 1850config PADATA 1851 depends on SMP 1852 bool 1853 1854# Can be selected by architectures with broken toolchains 1855# that get confused by correct const<->read_only section 1856# mappings 1857config BROKEN_RODATA 1858 bool 1859 1860config ASN1 1861 tristate 1862 help 1863 Build a simple ASN.1 grammar compiler that produces a bytecode output 1864 that can be interpreted by the ASN.1 stream decoder and used to 1865 inform it as to what tags are to be expected in a stream and what 1866 functions to call on what tags. 1867 1868source "kernel/Kconfig.locks" 1869