From a45a8c8571c0be6a6bd72ae5a14255c26b14b504 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 16 Sep 2011 21:24:02 +0200 Subject: hwmon: (coretemp) Let the user force TjMax On old CPUs (and even some recent Atom CPUs) TjMax can't be read from the CPU registers, so it is guessed by the driver using a complex heuristic which isn't reliable. So let users who know their CPU's TjMax pass it as a module parameter. Signed-off-by: Jean Delvare Cc: Fenghua Yu Cc: "R, Durgadoss" Cc: Guenter Roeck Cc: Alexander Stein Acked-by: Fenghua Yu Signed-off-by: Guenter Roeck --- Documentation/hwmon/coretemp | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'Documentation') diff --git a/Documentation/hwmon/coretemp b/Documentation/hwmon/coretemp index fa8776a..49c0d42 100644 --- a/Documentation/hwmon/coretemp +++ b/Documentation/hwmon/coretemp @@ -49,9 +49,10 @@ tempX_label - Contains string "Core X", where X is processor number. For Package temp, this will be "Physical id Y", where Y is the package number. -The TjMax temperature is set to 85 degrees C if undocumented model specific -register (UMSR) 0xee has bit 30 set. If not the TjMax is 100 degrees C as -(sometimes) documented in processor datasheet. +On CPU models which support it, TjMax is read from a model-specific register. +On other models, it is set to an arbitrary value based on weak heuristics. +If these heuristics don't work for you, you can pass the correct TjMax value +as a module parameter (tjmax). Appendix A. Known TjMax lists (TBD): Some information comes from ark.intel.com -- cgit v1.1 From f4af6fd6e21792ca4deca3d29c113a575594078e Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 19 Sep 2011 21:41:16 -0700 Subject: hwmon: (coretemp) Don't use threshold registers for tempX_max With commit c814a4c7c4aad795835583344353963a0a673eb0, the meaning of tempX_max was changed. It no longer returns the value of bits 8:15 of MSR_IA32_TEMPERATURE_TARGET, but instead returns the value of CPU threshold register T1. tempX_max_hyst was added to reflect the value of temperature threshold register T0. As it turns out, T0 and T1 are used on some systems, presumably by the BIOS. Also, T0 and T1 don't have a well defined meaning. The thresholds may be used as upper or lower limits, and it is not guaranteed that T0 <= T1. Thus, the new attribute mapping does not reflect the actual usage of the threshold registers. Also, register contents are changed during runtime by an entity other than the hwmon driver, meaning the values cached by the driver do not reflect actual register contents. Revert most of c814a4c7c4aad795835583344353963a0a673eb0 to address the problem. Support for T0 and T1 will be added back in with a separate commit, using new attribute names. Signed-off-by: Guenter Roeck Cc: Fenghua Yu Cc: Durgadoss R Acked-by: Jean Delvare --- Documentation/hwmon/coretemp | 7 ------- 1 file changed, 7 deletions(-) (limited to 'Documentation') diff --git a/Documentation/hwmon/coretemp b/Documentation/hwmon/coretemp index 49c0d42..84d46c0 100644 --- a/Documentation/hwmon/coretemp +++ b/Documentation/hwmon/coretemp @@ -35,13 +35,6 @@ the Out-Of-Spec bit. Following table summarizes the exported sysfs files: All Sysfs entries are named with their core_id (represented here by 'X'). tempX_input - Core temperature (in millidegrees Celsius). tempX_max - All cooling devices should be turned on (on Core2). - Initialized with IA32_THERM_INTERRUPT. When the CPU - temperature reaches this temperature, an interrupt is - generated and tempX_max_alarm is set. -tempX_max_hyst - If the CPU temperature falls below than temperature, - an interrupt is generated and tempX_max_alarm is reset. -tempX_max_alarm - Set if the temperature reaches or exceeds tempX_max. - Reset if the temperature drops to or below tempX_max_hyst. tempX_crit - Maximum junction temperature (in millidegrees Celsius). tempX_crit_alarm - Set when Out-of-spec bit is set, never clears. Correct CPU operation is no longer guaranteed. -- cgit v1.1 From e369fde1af20c7c67acff3bc4a36e451f38ae7d0 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 22 Sep 2011 14:11:38 -0700 Subject: thp: fix khugepaged defrag tunable documentation Commit e27e6151b154 ("mm/thp: use conventional format for boolean attributes") changed /sys/kernel/mm/transparent_hugepage/khugepaged/defrag to be tuned by using 1 (enabled) or 0 (disabled) instead of "yes" and "no", respectively. Update the documentation. Signed-off-by: David Rientjes Signed-off-by: Linus Torvalds --- Documentation/vm/transhuge.txt | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'Documentation') diff --git a/Documentation/vm/transhuge.txt b/Documentation/vm/transhuge.txt index 0924aac..29bdf62 100644 --- a/Documentation/vm/transhuge.txt +++ b/Documentation/vm/transhuge.txt @@ -123,10 +123,11 @@ be automatically shutdown if it's set to "never". khugepaged runs usually at low frequency so while one may not want to invoke defrag algorithms synchronously during the page faults, it should be worth invoking defrag at least in khugepaged. However it's -also possible to disable defrag in khugepaged: +also possible to disable defrag in khugepaged by writing 0 or enable +defrag in khugepaged by writing 1: -echo yes >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag -echo no >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag +echo 0 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag +echo 1 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag You can also control how many pages khugepaged should scan at each pass: -- cgit v1.1 From e451e61b56b0401442f7a306cd309c3b0c56c285 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Tue, 27 Sep 2011 13:26:27 -0400 Subject: net: fix a typo in Documentation/networking/scaling.txt Signed-off-by: Jason Wang Signed-off-by: David S. Miller --- Documentation/networking/scaling.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'Documentation') diff --git a/Documentation/networking/scaling.txt b/Documentation/networking/scaling.txt index 58fd741..8ce7c30 100644 --- a/Documentation/networking/scaling.txt +++ b/Documentation/networking/scaling.txt @@ -243,7 +243,7 @@ configured. The number of entries in the global flow table is set through: The number of entries in the per-queue flow table are set through: - /sys/class/net//queues/tx-/rps_flow_cnt + /sys/class/net//queues/rx-/rps_flow_cnt == Suggested Configuration -- cgit v1.1 From 605b91c8f619047ed6d88e24c531d4bdddad46f4 Mon Sep 17 00:00:00 2001 From: "Roy.Li" Date: Wed, 28 Sep 2011 19:51:54 +0000 Subject: net: Documentation: Fix type of variables Signed-off-by: Roy.Li Signed-off-by: David S. Miller --- Documentation/networking/ip-sysctl.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'Documentation') diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt index 8154699..ca5cdcd 100644 --- a/Documentation/networking/ip-sysctl.txt +++ b/Documentation/networking/ip-sysctl.txt @@ -1042,7 +1042,7 @@ conf/interface/*: The functional behaviour for certain settings is different depending on whether local forwarding is enabled or not. -accept_ra - BOOLEAN +accept_ra - INTEGER Accept Router Advertisements; autoconfigure using them. Possible values are: @@ -1106,7 +1106,7 @@ dad_transmits - INTEGER The amount of Duplicate Address Detection probes to send. Default: 1 -forwarding - BOOLEAN +forwarding - INTEGER Configure interface-specific Host/Router behaviour. Note: It is recommended to have the same setting on all -- cgit v1.1 From 186c6bbced722cfeff041d2a1264c95f5d042050 Mon Sep 17 00:00:00 2001 From: Benjamin Poirier Date: Tue, 4 Oct 2011 04:00:30 +0000 Subject: net: fix typos in Documentation/networking/scaling.txt The second hunk fixes rps_sock_flow_table but has to re-wrap the paragraph. Signed-off-by: Benjamin Poirier Signed-off-by: David S. Miller --- Documentation/networking/scaling.txt | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'Documentation') diff --git a/Documentation/networking/scaling.txt b/Documentation/networking/scaling.txt index 8ce7c30..fe67b5c 100644 --- a/Documentation/networking/scaling.txt +++ b/Documentation/networking/scaling.txt @@ -27,7 +27,7 @@ applying a filter to each packet that assigns it to one of a small number of logical flows. Packets for each flow are steered to a separate receive queue, which in turn can be processed by separate CPUs. This mechanism is generally known as “Receive-side Scaling” (RSS). The goal of RSS and -the other scaling techniques to increase performance uniformly. +the other scaling techniques is to increase performance uniformly. Multi-queue distribution can also be used for traffic prioritization, but that is not the focus of these techniques. @@ -186,10 +186,10 @@ are steered using plain RPS. Multiple table entries may point to the same CPU. Indeed, with many flows and few CPUs, it is very likely that a single application thread handles flows with many different flow hashes. -rps_sock_table is a global flow table that contains the *desired* CPU for -flows: the CPU that is currently processing the flow in userspace. Each -table value is a CPU index that is updated during calls to recvmsg and -sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() +rps_sock_flow_table is a global flow table that contains the *desired* CPU +for flows: the CPU that is currently processing the flow in userspace. +Each table value is a CPU index that is updated during calls to recvmsg +and sendmsg (specifically, inet_recvmsg(), inet_sendmsg(), inet_sendpage() and tcp_splice_read()). When the scheduler moves a thread to a new CPU while it has outstanding -- cgit v1.1