[tip: x86/urgent] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

The following commit has been merged into the x86/urgent branch of tip: Commit-ID: 06c5fe9b12dde1b62821f302f177c972bb1c81f9 Gitweb: https://git.kernel.org/tip/06c5fe9b12dde1b62821f302f177c972bb1c81f9 Author:Xiaochen Shen AuthorDate:Fri, 04 Dec 2020 14:27:59 +08:00

Re: [tip: x86/cache] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

Hi Boris, On 12/10/2020 18:26, Borislav Petkov wrote: On Thu, Dec 10, 2020 at 12:45:11PM +0800, Xiaochen Shen wrote: Thank you for clarifying this issue. It is not a 0-DAY CI issue. Which begs the question: this patch should be Cc: stable and should go in now, shouldn't it? Because the

[tip: x86/cache v2] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

local_1` Before fix: local b/w (bytes/s): 11076796416 After fix: local b/w (bytes/s): 5465014272 Fixes: ba0f26d8529c (x86/intel_rdt/mba_sc: Prepare for feedback loop) Signed-off-by: Xiaochen Shen Reviewed-by: Tony Luck --- arch/x86/kernel/cpu/resctrl/monitor.c | 6 ++ 1 file changed, 2

Re: [tip: x86/cache] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

Hi Boris, On 12/10/2020 6:23, Borislav Petkov wrote: And then it didn't apply cleanly: $ test-apply.sh /tmp/xiaochen.01 checking file arch/x86/kernel/cpu/resctrl/monitor.c Hunk #1 FAILED at 279. Hunk #2 succeeded at 514 (offset 64 lines). 1 out of 2 hunks FAILED I wiggled it in but it ended up

[tip: x86/cache] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

The following commit has been merged into the x86/cache branch of tip: Commit-ID: 2ba836dbe2467d31fffb439258c2f454c6f1a317 Gitweb: https://git.kernel.org/tip/2ba836dbe2467d31fffb439258c2f454c6f1a317 Author:Xiaochen Shen AuthorDate:Fri, 04 Dec 2020 14:27:59 +08:00

[PATCH] x86/resctrl: Fix incorrect local bandwidth when mba_sc is enabled

local_1` Before fix: local b/w (bytes/s): 11076796416 After fix: local b/w (bytes/s): 5465014272 Fixes: ba0f26d8529c (x86/intel_rdt/mba_sc: Prepare for feedback loop) Signed-off-by: Xiaochen Shen Reviewed-by: Tony Luck --- arch/x86/kernel/cpu/resctrl/monitor.c | 6 ++ 1 file changed, 2

[tip: x86/cache] x86/resctrl: Clean up unused function parameter in rmdir path

The following commit has been merged into the x86/cache branch of tip: Commit-ID: 19eb86a72df50adcf554f234469bb5b7209b7640 Gitweb: https://git.kernel.org/tip/19eb86a72df50adcf554f234469bb5b7209b7640 Author:Xiaochen Shen AuthorDate:Tue, 01 Dec 2020 02:06:58 +08:00

[PATCH v2] x86/resctrl: Clean up unused function parameter in rmdir path

the unused function parameter in rdtgroup_ctrl_remove(), rdtgroup_rmdir_mon() and their callers rdtgroup_rmdir_ctrl() and rdtgroup_rmdir(). Signed-off-by: Xiaochen Shen Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 17 +++-- 1 file changed, 7 insertions(+), 10

[tip: x86/urgent] x86/resctrl: Remove superfluous kernfs_get() calls to prevent refcount leak

The following commit has been merged into the x86/urgent branch of tip: Commit-ID: fd8d9db3559a29fd737bcdb7c4fcbe1940caae34 Gitweb: https://git.kernel.org/tip/fd8d9db3559a29fd737bcdb7c4fcbe1940caae34 Author:Xiaochen Shen AuthorDate:Sat, 31 Oct 2020 03:10:53 +08:00

[tip: x86/urgent] x86/resctrl: Add necessary kernfs_put() calls to prevent refcount leak

The following commit has been merged into the x86/urgent branch of tip: Commit-ID: 758999246965eeb8b253d47e72f7bfe508804b16 Gitweb: https://git.kernel.org/tip/758999246965eeb8b253d47e72f7bfe508804b16 Author:Xiaochen Shen AuthorDate:Sat, 31 Oct 2020 03:11:28 +08:00

Re: [PATCH 1/3] x86/resctrl: Remove superfluous kernfs_get() calls to prevent refcount leak

, 2020 at 03:03:58AM +0800, Xiaochen Shen wrote: Willem reported growing of kernfs_node_cache entries in slabtop when repeatedly creating and removing resctrl subdirectories as well as when repeatedly mounting and unmounting resctrl filesystem. On resource group (control as well as monitoring

Re: [PATCH 0/3] Fix kernfs node reference count leak issues

Hi Reinette, Thank you for correcting this! The subject of this "cover letter" should be: x86/resctrl: Fix kernfs node reference count leak issues On 10/31/2020 5:18, Reinette Chatre wrote: Apologies, the Subject intended to have a "x86/resctrl:" prefix. On 10/30/2020 12:

Re: [PATCH 1/3] x86/resctrl: Remove superfluous kernfs_get() calls to prevent refcount leak

Sorry, please ignore this patch. The correct version of patchset: https://lkml.org/lkml/2020/10/30/998 On 10/31/2020 3:03, Xiaochen Shen wrote: Willem reported growing of kernfs_node_cache entries in slabtop when repeatedly creating and removing resctrl subdirectories as well as when

[PATCH 3/3] x86/resctrl: Clean up unused function parameter in rmdir path

d function parameter in rdtgroup_ctrl_remove(), rdtgroup_rmdir_mon() and their callers rdtgroup_rmdir_ctrl() and rdtgroup_rmdir(). Signed-off-by: Xiaochen Shen Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 17 +++-- 1 file changed, 7 insertions(+), 10

[PATCH 2/3] x86/resctrl: Add necessary kernfs_put() calls to prevent refcount leak

illem de Bruijn Signed-off-by: Xiaochen Shen Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 32 +--- 1 file changed, 25 insertions(+), 7 deletions(-) diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kernel/cpu/resctrl

[PATCH 1/3] x86/resctrl: Remove superfluous kernfs_get() calls to prevent refcount leak

mkdir to resctrl file system") Fixes: 4e978d06dedb ("x86/intel_rdt: Add "info" files to resctrl file system") Reported-by: Willem de Bruijn Signed-off-by: Xiaochen Shen Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 35 ++---

[PATCH 1/3] x86/resctrl: Remove superfluous kernfs_get() calls to prevent refcount leak

mkdir to resctrl file system") Fixes: 4e978d06dedb ("x86/intel_rdt: Add "info" files to resctrl file system") Reported-by: Willem de Bruijn Signed-off-by: Xiaochen Shen Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 35 ++---

[PATCH 0/3] Fix kernfs node reference count leak issues

Fix several kernfs node reference count leak issues: (1) Remove superfluous kernfs_get() calls to prevent refcount leak (2) Add necessary kernfs_put() calls to prevent refcount leak (3) Follow-up cleanup for the change in previous patch. Xiaochen Shen (3): x86/resctrl: Remove superfluous

Re: [tip:x86/cache] x86/resctrl: Initialize a new resource group with default MBA values

Hi Boris, On 4/18/2019 15:28, Borislav Petkov wrote: On Thu, Apr 18, 2019 at 03:03:35PM +0800, Xiaochen Shen wrote: In my opinion, this newline is unnecessary. Thank you. See commit message: [ bp: Add newlines between code blocks for better readability. ] I got this commit message

Re: [tip:x86/cache] x86/resctrl: Initialize a new resource group with default MBA values

Hi Boris, I found a nitpick - an unnecessary newline at the end of the patch. Please help double check. Thank you. On 4/18/2019 6:14, tip-bot for Xiaochen Shen wrote: Commit-ID: 47820e73f5b3a1fdb8ebd1219191edc96e0c85c1 Gitweb: https://git.kernel.org/tip

[tip:x86/cache] x86/resctrl: Initialize a new resource group with default MBA values

Commit-ID: 47820e73f5b3a1fdb8ebd1219191edc96e0c85c1 Gitweb: https://git.kernel.org/tip/47820e73f5b3a1fdb8ebd1219191edc96e0c85c1 Author: Xiaochen Shen AuthorDate: Wed, 17 Apr 2019 19:08:49 +0800 Committer: Borislav Petkov CommitDate: Thu, 18 Apr 2019 00:06:31 +0200 x86/resctrl

[tip:x86/cache] x86/resctrl: Move per RDT domain initialization to a separate function

Commit-ID: 7390619ab9ea9fd0ba9f4c3e4749ee20262cba7d Gitweb: https://git.kernel.org/tip/7390619ab9ea9fd0ba9f4c3e4749ee20262cba7d Author: Xiaochen Shen AuthorDate: Wed, 17 Apr 2019 19:08:48 +0800 Committer: Borislav Petkov CommitDate: Wed, 17 Apr 2019 23:59:56 +0200 x86/resctrl: Move

[tip:x86/cache] x86/resctrl: Initialize a new resource group with default MBA values

Commit-ID: 7b05c9c1fd39cb08017353a47e7ff14ad1a3b875 Gitweb: https://git.kernel.org/tip/7b05c9c1fd39cb08017353a47e7ff14ad1a3b875 Author: Xiaochen Shen AuthorDate: Wed, 17 Apr 2019 19:08:49 +0800 Committer: Borislav Petkov CommitDate: Wed, 17 Apr 2019 21:18:05 +0200 x86/resctrl

[tip:x86/cache] x86/resctrl: Move per RDT domain initialization to a separate function

Commit-ID: 8cea8808525eb3a36f3d54412843948f169d6a6e Gitweb: https://git.kernel.org/tip/8cea8808525eb3a36f3d54412843948f169d6a6e Author: Xiaochen Shen AuthorDate: Wed, 17 Apr 2019 19:08:48 +0800 Committer: Borislav Petkov CommitDate: Wed, 17 Apr 2019 21:03:38 +0200 x86/resctrl: Move

[PATCH v2 2/2] x86/resctrl: Initialize a new resource group with default MBA values

rce and cache resources in separate functions. Signed-off-by: Xiaochen Shen Reviewed-by: Fenghua Yu Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 4 +-- arch/x86/kernel/cpu/resctrl/rdtgroup.c| 52 +++ 2 files changed, 34 insertions(+),

[PATCH v2 0/2] Initialize a new resource group with default MBA values

. Xiaochen Shen (2): x86/resctrl: Move per RDT domain initialization to a separate function x86/resctrl: Initialize a new resource group with default MBA values arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 4 +- arch/x86/kernel/cpu/resctrl/rdtgroup.c| 171

[PATCH v2 1/2] x86/resctrl: Move per RDT domain initialization to a separate function

Carve out per rdt_domain initialization code in rdtgroup_init_alloc() into a separate function. No functional change, this preparatory patch will make the code more readable and save us at least two indentation levels. Signed-off-by: Xiaochen Shen --- arch/x86/kernel/cpu/resctrl/rdtgroup.c

Re: [PATCH 2/2] x86/resctrl: Initialize new resource group with default MBA values

Hi Boris, On 4/17/2019 4:59, Borislav Petkov wrote: I guess but be conservative when adding too many words to a function's name - that might get heavy when reading the code later. For example, there's an argument to not suffix it with "_cat": it is only called by rdtgroup_init_cat() - rdtgroup_

Re: [PATCH 2/2] x86/resctrl: Initialize new resource group with default MBA values

Hi Boris, Thank you very much for code review. I will fix these issues in v2 patch. Please find more comments inline. Thank you. On 4/15/2019 19:34, Borislav Petkov wrote: On Wed, Apr 10, 2019 at 03:55:28AM +0800, Xiaochen Shen wrote: Currently when a new resource group is created, the

[tip:x86/urgent] x86/resctrl: Do not repeat rdtgroup mode initialization

Commit-ID: 40fba00ffa431c8597ca785ea1cfa4d9f6503390 Gitweb: https://git.kernel.org/tip/40fba00ffa431c8597ca785ea1cfa4d9f6503390 Author: Xiaochen Shen AuthorDate: Wed, 10 Apr 2019 03:53:49 +0800 Committer: Borislav Petkov CommitDate: Sun, 14 Apr 2019 18:23:13 +0200 x86/resctrl: Do not

[PATCH 2/2] x86/resctrl: Initialize new resource group with default MBA values

Signed-off-by: Xiaochen Shen Reviewed-by: Fenghua Yu Reviewed-by: Reinette Chatre --- arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 4 +- arch/x86/kernel/cpu/resctrl/rdtgroup.c| 139 -- 2 files changed, 75 insertions(+), 68 deletions(-) diff --git a/arch/x86/kerne

[PATCH 1/2] x86/resctrl: Fix repeating rdtgroup mode initialization

Initialize new resource group with sane defaults") Signed-off-by: Xiaochen Shen Acked-by: Reinette Chatre Reviewed-by: Fenghua Yu --- arch/x86/kernel/cpu/resctrl/rdtgroup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kern

[tip:x86/urgent] x86/resctrl: Fix typos in the mba_sc mount option

Commit-ID: faa3604eda325588451c9c1eb4c8a8d04c1cd633 Gitweb: https://git.kernel.org/tip/faa3604eda325588451c9c1eb4c8a8d04c1cd633 Author: Xiaochen Shen AuthorDate: Sat, 30 Mar 2019 05:50:38 +0800 Committer: Borislav Petkov CommitDate: Mon, 1 Apr 2019 18:45:27 +0200 x86/resctrl: Fix

[PATCH] x86/resctrl: Fix typos in mba_sc mount option

;) changes the mount option from "mba_MBps" to "mba_mpbs" by mistake. Change back from "mba_mpbs" to "mba_MBps", and rename Opt_mba_mpbs to Opt_mba_mbps. Fixes: 23bf1b6be9c2 ("kernfs, sysfs, cgroup, intel_rdt: Support fs_context") Signed-off-by: Xiaoche

[tip:x86/cache] x86/intel_rdt: Show missing resctrl mount options

Commit-ID: 2cc81c6992248ea37d0241bc325977bab310bc3b Gitweb: https://git.kernel.org/tip/2cc81c6992248ea37d0241bc325977bab310bc3b Author: Xiaochen Shen AuthorDate: Wed, 12 Sep 2018 16:48:38 -0700 Committer: Thomas Gleixner CommitDate: Wed, 3 Oct 2018 21:53:49 +0200 x86/intel_rdt: Show

[tip:x86/cache] x86/intel_rdt: Show missing resctrl mount options

Commit-ID: 32ece95d3052e0778eab98fd9c2ffde92c599ce9 Gitweb: https://git.kernel.org/tip/32ece95d3052e0778eab98fd9c2ffde92c599ce9 Author: Xiaochen Shen AuthorDate: Wed, 12 Sep 2018 16:48:38 -0700 Committer: Thomas Gleixner CommitDate: Wed, 3 Oct 2018 16:19:36 +0200 x86/intel_rdt: Show

[tip:x86/urgent] x86/intel_rdt: Fix data type in parsing callbacks

Commit-ID: 753694a8df318f204a0ac1303de136def16f2e9c Gitweb: https://git.kernel.org/tip/753694a8df318f204a0ac1303de136def16f2e9c Author: Xiaochen Shen AuthorDate: Sat, 15 Sep 2018 14:58:19 -0700 Committer: Thomas Gleixner CommitDate: Tue, 18 Sep 2018 23:38:05 +0200 x86/intel_rdt: Fix

[tip:x86/cache] x86/intel_rdt: Fix a silent failure when writing zero value schemata

Commit-ID: 2244645ab194fe45ffcbaa08f235c8f0c7fb54fc Gitweb: https://git.kernel.org/tip/2244645ab194fe45ffcbaa08f235c8f0c7fb54fc Author: Xiaochen Shen AuthorDate: Fri, 10 Nov 2017 11:16:24 -0800 Committer: Thomas Gleixner CommitDate: Sun, 12 Nov 2017 09:01:40 +0100 x86/intel_rdt: Fix a

[tip:perf/core] perf tests: Remove Intel CQM perf test

Commit-ID: 5c9295bfe6f5f59f3f2eee78f58b0523d117897e Gitweb: http://git.kernel.org/tip/5c9295bfe6f5f59f3f2eee78f58b0523d117897e Author: Xiaochen Shen AuthorDate: Tue, 19 Sep 2017 12:57:37 +0800 Committer: Arnaldo Carvalho de Melo CommitDate: Thu, 21 Sep 2017 13:12:58 -0300 perf tests

[PATCH] perf tests: Remove Intel CQM perf test

Intel CQM perf test is obsolete for perf PMU code has been removed in commit c39a0e2c8850 ("x86/perf/cqm: Wipe out perf based cqm"). Signed-off-by: Xiaochen Shen --- tools/perf/arch/x86/include/arch-tests.h | 1 - tools/perf/arch/x86/tests/Build | 1 - tools/perf/arch

[tip:x86/urgent] x86/intel_rdt: Fix a typo in Documentation

Commit-ID: fb8fb46c56289b3f34b5d90a4ec65e9e4e4544a5 Gitweb: http://git.kernel.org/tip/fb8fb46c56289b3f34b5d90a4ec65e9e4e4544a5 Author: Xiaochen Shen AuthorDate: Wed, 3 May 2017 11:15:56 +0800 Committer: Thomas Gleixner CommitDate: Tue, 9 May 2017 09:41:42 +0200 x86/intel_rdt: Fix a

Re: [PATCH] x86/intel_rdt: Fix a typo in Documentation

Thank you, Thomas. I will send out v2 patch with the fix soon. Best regards, Xiaochen On 2017/5/9 15:22, Thomas Gleixner wrote: > On Tue, 9 May 2017, Xiaochen Shen wrote: > >> Hi Thomas, >> >> I am sorry, this is the reason why I added Fenghua in SOB list: >> I h

Re: [PATCH] x86/intel_rdt: Fix a typo in Documentation

n On 2017/5/9 15:03, Thomas Gleixner wrote: > On Wed, 3 May 2017, Xiaochen Shen wrote: > >> The typo is in example 3. >> >> "C0" in "# echo C0 > p0/cpus" is wrong because it specifies core >> 6-7 instead of wanted core 4-7. >> >> Co

[PATCH] x86/intel_rdt: Fix a typo in Documentation

The typo is in example 3. "C0" in "# echo C0 > p0/cpus" is wrong because it specifies core 6-7 instead of wanted core 4-7. Correct this typo to avoid confusion. Signed-off-by: Xiaochen Shen Signed-off-by: Fenghua Yu --- Documentation/x86/intel_rdt_ui.txt | 2 +- 1 fi

[PATCH] x86/intel_rdt: Fix two typos in Documentation

Both typos are in example 3. Because cache id 0 is the only cache id, the ";" is redundant in "# echo "L3:0=ffc00;" > p0/schemata". And "C0" in "# echo C0 > p0/cpus" is wrong because it specifies core 6-7 instead of wanted core 4-7. Correc