aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation/kernel-parameters.txt
diff options
context:
space:
mode:
Diffstat (limited to 'Documentation/kernel-parameters.txt')
-rw-r--r--Documentation/kernel-parameters.txt81
1 files changed, 77 insertions, 4 deletions
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 479f33204a37..4df73da11adc 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -829,6 +829,15 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
829 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 829 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
830 tracking down these problems. 830 tracking down these problems.
831 831
832 debug_pagealloc=
833 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
834 parameter enables the feature at boot time. In
835 default, it is disabled. We can avoid allocating huge
836 chunk of memory for debug pagealloc if we don't enable
837 it at boot time and the system will work mostly same
838 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
839 on: enable the feature
840
832 debugpat [X86] Enable PAT debugging 841 debugpat [X86] Enable PAT debugging
833 842
834 decnet.addr= [HW,NET] 843 decnet.addr= [HW,NET]
@@ -1228,9 +1237,7 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
1228 multiple times interleaved with hugepages= to reserve 1237 multiple times interleaved with hugepages= to reserve
1229 huge pages of different sizes. Valid pages sizes on 1238 huge pages of different sizes. Valid pages sizes on
1230 x86-64 are 2M (when the CPU supports "pse") and 1G 1239 x86-64 are 2M (when the CPU supports "pse") and 1G
1231 (when the CPU supports the "pdpe1gb" cpuinfo flag) 1240 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1232 Note that 1GB pages can only be allocated at boot time
1233 using hugepages= and not freed afterwards.
1234 1241
1235 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1242 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1236 terminal devices. Valid values: 0..8 1243 terminal devices. Valid values: 0..8
@@ -1369,6 +1376,10 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
1369 Formats: { "ima" | "ima-ng" } 1376 Formats: { "ima" | "ima-ng" }
1370 Default: "ima-ng" 1377 Default: "ima-ng"
1371 1378
1379 ima_template_fmt=
1380 [IMA] Define a custom template format.
1381 Format: { "field1|...|fieldN" }
1382
1372 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1383 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1373 Format: <min_file_size> 1384 Format: <min_file_size>
1374 Set the minimal file size for using asynchronous hash. 1385 Set the minimal file size for using asynchronous hash.
@@ -1446,6 +1457,18 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
1446 disable 1457 disable
1447 Do not enable intel_pstate as the default 1458 Do not enable intel_pstate as the default
1448 scaling driver for the supported processors 1459 scaling driver for the supported processors
1460 force
1461 Enable intel_pstate on systems that prohibit it by default
1462 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1463 instead of acpi-cpufreq may disable platform features, such
1464 as thermal controls and power capping, that rely on ACPI
1465 P-States information being indicated to OSPM and therefore
1466 should be used with caution. This option does not work with
1467 processors that aren't supported by the intel_pstate driver
1468 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1469 no_hwp
1470 Do not enable hardware P state control (HWP)
1471 if available.
1449 1472
1450 intremap= [X86-64, Intel-IOMMU] 1473 intremap= [X86-64, Intel-IOMMU]
1451 on enable Interrupt Remapping (default) 1474 on enable Interrupt Remapping (default)
@@ -2503,12 +2526,21 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
2503 OSS [HW,OSS] 2526 OSS [HW,OSS]
2504 See Documentation/sound/oss/oss-parameters.txt 2527 See Documentation/sound/oss/oss-parameters.txt
2505 2528
2529 page_owner= [KNL] Boot-time page_owner enabling option.
2530 Storage of the information about who allocated
2531 each page is disabled in default. With this switch,
2532 we can turn it on.
2533 on: enable the feature
2534
2506 panic= [KNL] Kernel behaviour on panic: delay <timeout> 2535 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2507 timeout > 0: seconds before rebooting 2536 timeout > 0: seconds before rebooting
2508 timeout = 0: wait forever 2537 timeout = 0: wait forever
2509 timeout < 0: reboot immediately 2538 timeout < 0: reboot immediately
2510 Format: <timeout> 2539 Format: <timeout>
2511 2540
2541 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2542 on a WARN().
2543
2512 crash_kexec_post_notifiers 2544 crash_kexec_post_notifiers
2513 Run kdump after running panic-notifiers and dumping 2545 Run kdump after running panic-notifiers and dumping
2514 kmsg. This only for the users who doubt kdump always 2546 kmsg. This only for the users who doubt kdump always
@@ -2940,6 +2972,13 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
2940 quiescent states. Units are jiffies, minimum 2972 quiescent states. Units are jiffies, minimum
2941 value is one, and maximum value is HZ. 2973 value is one, and maximum value is HZ.
2942 2974
2975 rcutree.kthread_prio= [KNL,BOOT]
2976 Set the SCHED_FIFO priority of the RCU
2977 per-CPU kthreads (rcuc/N). This value is also
2978 used for the priority of the RCU boost threads
2979 (rcub/N). Valid values are 1-99 and the default
2980 is 1 (the least-favored priority).
2981
2943 rcutree.rcu_nocb_leader_stride= [KNL] 2982 rcutree.rcu_nocb_leader_stride= [KNL]
2944 Set the number of NOCB kthread groups, which 2983 Set the number of NOCB kthread groups, which
2945 defaults to the square root of the number of 2984 defaults to the square root of the number of
@@ -3089,6 +3128,15 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
3089 messages. Disable with a value less than or equal 3128 messages. Disable with a value less than or equal
3090 to zero. 3129 to zero.
3091 3130
3131 rcupdate.rcu_self_test= [KNL]
3132 Run the RCU early boot self tests
3133
3134 rcupdate.rcu_self_test_bh= [KNL]
3135 Run the RCU bh early boot self tests
3136
3137 rcupdate.rcu_self_test_sched= [KNL]
3138 Run the RCU sched early boot self tests
3139
3092 rdinit= [KNL] 3140 rdinit= [KNL]
3093 Format: <full_path> 3141 Format: <full_path>
3094 Run specified binary instead of /init from the ramdisk, 3142 Run specified binary instead of /init from the ramdisk,
@@ -3412,6 +3460,13 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
3412 neutralize any effect of /proc/sys/kernel/sysrq. 3460 neutralize any effect of /proc/sys/kernel/sysrq.
3413 Useful for debugging. 3461 Useful for debugging.
3414 3462
3463 tcpmhash_entries= [KNL,NET]
3464 Set the number of tcp_metrics_hash slots.
3465 Default value is 8192 or 16384 depending on total
3466 ram pages. This is used to specify the TCP metrics
3467 cache size. See Documentation/networking/ip-sysctl.txt
3468 "tcp_no_metrics_save" section for more details.
3469
3415 tdfx= [HW,DRM] 3470 tdfx= [HW,DRM]
3416 3471
3417 test_suspend= [SUSPEND][,N] 3472 test_suspend= [SUSPEND][,N]
@@ -3501,7 +3556,7 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
3501 are saved. 3556 are saved.
3502 3557
3503 trace_buf_size=nn[KMG] 3558 trace_buf_size=nn[KMG]
3504 [FTRACE] will set tracing buffer size. 3559 [FTRACE] will set tracing buffer size on each cpu.
3505 3560
3506 trace_event=[event-list] 3561 trace_event=[event-list]
3507 [FTRACE] Set and start specified trace events in order 3562 [FTRACE] Set and start specified trace events in order
@@ -3524,6 +3579,24 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
3524 See also Documentation/trace/ftrace.txt "trace options" 3579 See also Documentation/trace/ftrace.txt "trace options"
3525 section. 3580 section.
3526 3581
3582 tp_printk[FTRACE]
3583 Have the tracepoints sent to printk as well as the
3584 tracing ring buffer. This is useful for early boot up
3585 where the system hangs or reboots and does not give the
3586 option for reading the tracing buffer or performing a
3587 ftrace_dump_on_oops.
3588
3589 To turn off having tracepoints sent to printk,
3590 echo 0 > /proc/sys/kernel/tracepoint_printk
3591 Note, echoing 1 into this file without the
3592 tracepoint_printk kernel cmdline option has no effect.
3593
3594 ** CAUTION **
3595
3596 Having tracepoints sent to printk() and activating high
3597 frequency tracepoints such as irq or sched, can cause
3598 the system to live lock.
3599
3527 traceoff_on_warning 3600 traceoff_on_warning
3528 [FTRACE] enable this option to disable tracing when a 3601 [FTRACE] enable this option to disable tracing when a
3529 warning is hit. This turns off "tracing_on". Tracing can 3602 warning is hit. This turns off "tracing_on". Tracing can