CCF_100/linux-tkg
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Introduce linux58-rc-tkg

Browse Source
This commit is contained in:
Tk-Glitch 2020-06-15 03:32:10 +02:00
parent 7304411cbd
commit 687f886abf
18 changed files with 24283 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1120
linux58-rc-tkg/PKGBUILD Normal file
View File

File diff suppressed because it is too large Load Diff

36
linux58-rc-tkg/README.md Normal file
View File

@ -0,0 +1,36 @@
**Due to intel_pstate poor performances as of late, I have decided to set it to passive mode to make use of the acpi_cpufreq governors passthrough, keeping full support for turbo frequencies.**
### MuQSS and BMQ are not yet available options for this revision
### Nvidia proprietary drivers are not yet compatible with this revision
A custom Linux kernel 5.8 RC with specific PDS and CFS CPU schedulers related patchsets selector and added tweaks for a nice interactivity/performance balance, aiming for the best gaming experience.
PDS-mq was originally created by Alfred Chen : http://cchalpha.blogspot.com/
While he dropped it with kernel 5.1 in favor of its BMQ evolution/rework, my pretty bad gaming experiences with BMQ up to this point convinced me to keep PDS afloat for as long as it'll make sense/I'll be able to.
Various personalization options available and userpatches support (put your own patches in the same dir as the PKGBUILD, with the ".mypatch" extension.
Comes with a slightly modified Arch config asking for a few core personalization settings at compilation time.
If you want to streamline your kernel config for lower footprint and faster compilations : https://wiki.archlinux.org/index.php/Modprobed-db
You can enable support for it at the beginning of the PKGBUILD file. Make sure to read everything you need to know about it.
## Other stuff included:
- Per-CPU-arch native optimizations
- memory management and swapping tweaks
- scheduling tweaks
- optional "Zenify" patchset using core blk, mm and scheduler tweaks from Zen
- CFS tweaks
- using yeah TCP congestion algo by default
- using cake network queue management system
- using vm.max_map_count=262144 by default
- cherry-picked clear linux patches
- **optional** overrides for missing ACS capabilities
- **optional** Fsync support (proton)
```
git clone https://github.com/Frogging-Family/linux-tkg.git
cd linux-tkg/linux58-rc-tkg
makepkg -si
```

172
linux58-rc-tkg/customization.cfg Normal file
View File

@ -0,0 +1,172 @@
# linux58-TkG config file
#### MISC OPTIONS ####
# External config file to use - If the given file exists in path, it will override default config (customization.cfg) - Default is ~/.config/frogminer/linux52-tkg.cfg
_EXT_CONFIG_PATH=~/.config/frogminer/linux58-tkg.cfg
# Set to anything else than "true" to limit cleanup operations and keep source and files generated during compilation.
# Default is "true".
_NUKR="true"
# Custom compiler root dirs - Leave empty to use system compilers
# Example: CUSTOM_GCC_PATH="/home/frog/PKGBUILDS/mostlyportable-gcc/gcc-mostlyportable-9.2.0"
CUSTOM_GCC_PATH=""
# Set to the number corresponding to a predefined profile to use it. Current list of available profiles :
# 1 - Custom (meaning nothing will be enforced and you get to configure everything)
# 2 - Ryzen desktop (performance)
# 3 - Generic Desktop (Performance)
_OPTIPROFILE=""
# Set to true to bypass makepkg.conf and use all available threads for compilation. False will respect your makepkg.conf options.
_force_all_threads="true"
# Set to true to use modprobed db to clean config from unneeded modules. Speeds up compilation considerably. Requires root - https://wiki.archlinux.org/index.php/Modprobed-db
# !!!! Make sure to have a well populated db !!!! - Leave empty to be asked about it at build time
_modprobeddb="false"
# Set to "1" to call make menuconfig or "2" to call make nconfig before building the kernel. Set to false to disable and skip the prompt.
_menunconfig=""
# Set to true to generate a kernel config fragment from your changes in menuconfig/nconfig. Set to false to disable and skip the prompt.
_diffconfig=""
# Set to the file name where the generated config fragment should be written to. Only used if _diffconfig is active.
_diffconfig_name=""
#### KERNEL OPTIONS ####
# Name of the default config file to use from the linux???-tkg-config folder. Arch default is "config.x86_64" and Arch hardened is "config_hardened.x86_64".
# To get a complete hardened setup, you have to use "cfs" as _cpusched
_configfile="config.x86_64"
# Disable some non-module debugging - See PKGBUILD for the list
_debugdisable="false"
# LEAVE AN EMPTY VALUE TO BE PROMPTED ABOUT FOLLOWING OPTIONS AT BUILD TIME
# CPU scheduler - Options are "pds" or "cfs"
# "pds" is the recommended option for gaming
_cpusched=""
# CPU sched_yield_type - Choose what sort of yield sched_yield will perform
# For PDS and MuQSS: 0: No yield. (Recommended option for gaming on PDS and MuQSS)
# 1: Yield only to better priority/deadline tasks. (Default - can be unstable with PDS on some platforms)
# 2: Expire timeslice and recalculate deadline. (Usually the slowest option for PDS and MuQSS, not recommended)
# For BMQ: 0: No yield.
# 1: Deboost and requeue task. (Default)
# 2: Set rq skip task.
_sched_yield_type="0"
# Round Robin interval is the longest duration two tasks with the same nice level will be delayed for. When CPU time is requested by a task, it receives a time slice equal
# to the rr_interval in addition to a virtual deadline. When using yield_type 2, a low value can help offset the disadvantages of rescheduling a process that has yielded.
# MuQSS default: 6ms"
# PDS default: 4ms"
# BMQ default: 4ms"
# Set to "1" for 2ms, "2" for 4ms, "3" for 6ms, "4" for 8ms, or "default" to keep the chosen scheduler defaults.
_rr_interval=""
# Set to "true" to disable FUNCTION_TRACER/GRAPH_TRACER, lowering overhead but limiting debugging and analyzing of kernel functions - Kernel default is "false"
_ftracedisable="false"
# Set to "true" to disable NUMA, lowering overhead, but breaking CUDA/NvEnc on Nvidia equipped systems - Kernel default is "false"
_numadisable="false"
# Set to "1" to use CattaRappa mode (enabling full tickless), "2" for tickless idle only, or "0" for periodic ticks.
# Full tickless can give higher performances in various cases but, depending on hardware, lower consistency. Just tickless idle can perform better on some platforms (mostly AMD based).
_tickless=""
# Setting this to to "true" can improve latency on PDS (at the cost of throughput) and improve throughput on other schedulers (at the cost of latency) - Can improve VMs performance - Kernel default is "false"
_voluntary_preempt=""
# Set to "true" to enable Device Tree and Open Firmware support. If you don't know about it, you don't need it - Default is "false"
_OFenable="false"
# Set to "true" to use ACS override patch - https://wiki.archlinux.org/index.php/PCI_passthrough_via_OVMF#Bypassing_the_IOMMU_groups_.28ACS_override_patch.29 - Kernel default is "false"
_acs_override=""
# Set to "true" to add back missing symbol for AES-NI/AVX support on ZFS - https://github.com/NixOS/nixpkgs/blob/master/pkgs/os-specific/linux/kernel/export_kernel_fpu_functions.patch - Kernel default is "false"
_zfsfix="true"
# Set to "true" to enable support for fsync, an experimental replacement for esync found in Valve Proton 4.11+ - https://steamcommunity.com/games/221410/announcements/detail/2957094910196249305
_fsync=""
# A selection of patches from Zen/Liquorix kernel and additional tweaks for a better gaming experience (ZENIFY) - Default is "true"
_zenify="true"
# compiler optimization level - 1. Optimize for performance (-O2); 2. Optimize harder (-O3); 3. Optimize for size (-Os) - Kernel default is "1"
_compileroptlevel="1"
# CPU compiler optimizations - Defaults to generic optimizations if left empty
# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2"
# Intel CPUs : "mpsc"(P4 & older Netburst based Xeon) "atom" "core2" "nehalem" "westmere" "silvermont" "sandybridge" "ivybridge" "haswell" "broadwell" "skylake" "skylakex" "cannonlake" "icelake" "goldmont" "goldmontplus" "cascadelake"
# Other options :
# - "generic" (to share the package between machines with different CPUs)
# - "native" (use compiler autodetection and will prompt for P6_NOPS - Selecting your arch manually in the list above is recommended instead of this option)
_processor_opt=""
# MuQSS only - Make IRQ threading compulsory (FORCE_IRQ_THREADING) - Default is "false"
_irq_threading="false"
# MuQSS and PDS only - SMT (Hyperthreading) aware nice priority and policy support (SMT_NICE) - Kernel default is "true" - You can disable this on non-SMT/HT CPUs for lower overhead
_smt_nice=""
# Trust the CPU manufacturer to initialize Linux's CRNG (RANDOM_TRUST_CPU) - Kernel default is "false"
_random_trust_cpu="false"
# MuQSS only - CPU scheduler runqueue sharing - No sharing (RQ_NONE), SMT (hyperthread) siblings (RQ_SMT), Multicore siblings (RQ_MC), Symmetric Multi-Processing (RQ_SMP), NUMA (RQ_ALL)
# Valid values are "none", "smt", "mc", "mc-llc"(for zen), "smp", "all" - Kernel default is "smt"
_runqueue_sharing=""
# Timer frequency - "100" "500", "750" or "1000" - More options available in kernel config prompt when left empty depending on selected cpusched - Kernel default is "500" - For MuQSS, 100Hz is recommended
_timer_freq=""
# Default CPU governor - "performance", "ondemand", "schedutil" or leave empty for default (schedutil)
_default_cpu_gov="ondemand"
# Use an aggressive ondemand governor instead of default ondemand to improve performance on low loads/high core count CPUs while keeping some power efficiency from frequency scaling.
# It still requires you to either set ondemand as default governor or to select it some way.
_aggressive_ondemand="true"
# On some platforms, an acpi_cpufreq bug affects performance negatively. Set to "true" to disable it as a workaround, but it will use more power.
# https://github.com/Tk-Glitch/PKGBUILDS/issues/263
_disable_acpi_cpufreq=""
# You can pass a default set of kernel command line options here - example: "intel_pstate=passive nowatchdog amdgpu.ppfeaturemask=0xfffd7fff mitigations=off"
_custom_commandline="intel_pstate=passive"
#### SPESHUL OPTION ####
# If you want to bypass the stock naming scheme and enforce something else (example : "linux") - Useful for some bootloaders requiring manual entry editing on each release.
# !!! It will also change pkgname - If you don't explicitely need this, don't use it !!!
_custom_pkgbase=""
#### USER PATCHES ####
# community patches - add patches (separated by a space) of your choice by name from the community-patches dir
# example: _community_patches="clear_nack_in_tend_isr.myrevert ffb_regression_fix.mypatch 0008-drm-amd-powerplay-force-the-trim-of-the-mclk-dpm-levels-if-OD-is-enabled.mypatch"
_community_patches=""
# You can use your own patches by putting them in the same folder as the PKGBUILD and giving them the .mypatch extension.
# You can also revert patches by putting them in the same folder as the PKGBUILD and giving them the .myrevert extension.
# Also, userpatches variable below must be set to true for the above to work.
_user_patches="true"
# Apply all user patches without confirmation - !!! NOT RECOMMENDED !!!
_user_patches_no_confirm="false"
#### CONFIG FRAGMENTS ####
# You can use your own kernel config fragments by putting them in the same folder as the PKGBUILD and giving them the .myfrag extension.
# Also, the config fragments variable below must be set to true for the above to work.
_config_fragments="true"
# Apply all config fragments without confirmation - !!! NOT RECOMMENDED !!!
_config_fragments_no_confirm="false"

14
linux58-rc-tkg/linux58-tkg-config/90-cleanup.hook Normal file
View File

@ -0,0 +1,14 @@
[Trigger]
Type = File
Operation = Install
Operation = Upgrade
Operation = Remove
Target = usr/lib/modules/*/
Target = !usr/lib/modules/*/?*
[Action]
Description = Cleaning up...
When = PostTransaction
Exec = /usr/share/libalpm/scripts/cleanup
NeedsTargets

10
linux58-rc-tkg/linux58-tkg-config/cleanup Executable file
View File

@ -0,0 +1,10 @@
#!/bin/bash
for _f in /usr/lib/modules/*tkg*; do
if [[ ! -e ${_f}/vmlinuz ]]; then
rm -rf "$_f"
fi
done
# vim:set ft=sh sw=2 et:

10920
linux58-rc-tkg/linux58-tkg-config/config.x86_64 Normal file
View File

File diff suppressed because it is too large Load Diff

55
linux58-rc-tkg/linux58-tkg-config/generic-desktop-profile.cfg Normal file
View File

@ -0,0 +1,55 @@
# linux58-TkG config file
# Generic Desktop
#### MISC OPTIONS ####
# External config file to use - If the given file exists in path, it will override default config (customization.cfg) - Default is ~/.config/frogminer/linux50-tkg.cfg
_EXT_CONFIG_PATH=~/.config/frogminer/linux58-tkg.cfg
#### KERNEL OPTIONS ####
# Name of the default config file to use from the linux???-tkg-config folder. Arch default is "config.x86_64".
_configfile="config.x86_64"
# Disable some non-module debugging - See PKGBUILD for the list
_debugdisable="false"
# LEAVE AN EMPTY VALUE TO BE PROMPTED ABOUT FOLLOWING OPTIONS AT BUILD TIME
# Set to "true" to disable FUNCTION_TRACER/GRAPH_TRACER, lowering overhead but limiting debugging and analyzing of kernel functions - Kernel default is "false"
_ftracedisable="false"
# Set to "true" to disable NUMA, lowering overhead, but breaking CUDA/NvEnc on Nvidia equipped systems - Kernel default is "false"
_numadisable="false"
# Set to "true" to use explicit preemption points to lower latency at the cost of a small throughput loss - Can give a nice perf boost in VMs - Kernel default is "false"
_voluntary_preempt="false"
# A selection of patches from Zen/Liquorix kernel and additional tweaks for a better gaming experience (ZENIFY) - Default is "true"
_zenify="true"
# compiler optimization level - 1. Optimize for performance (-O2); 2. Optimize harder (-O3); 3. Optimize for size (-Os) - Kernel default is "2"
_compileroptlevel="1"
# Trust the CPU manufacturer to initialize Linux's CRNG (RANDOM_TRUST_CPU) - Kernel default is "false"
_random_trust_cpu="false"
# CPU scheduler runqueue sharing - No sharing (RQ_NONE), SMT (hyperthread) siblings (RQ_SMT), Multicore siblings (RQ_MC), Symmetric Multi-Processing (RQ_SMP), NUMA (RQ_ALL)
# Valid values are "none", "smt", "mc", "mc-llc"(for zen), "smp", "all" - Kernel default is "mc"
_runqueue_sharing="mc"
# Timer frequency - "500", "750" or "1000" - More options available in kernel config prompt when left empty depending on selected cpusched - Kernel default is "750"
_timer_freq="500"
#### USER PATCHES ####
# You can use your own patches by putting them in the same folder as the PKGBUILD and giving them the .mypatch extension.
# You can also revert patches by putting them in the same folder as the PKGBUILD and giving them the .myrevert extension.
# Also, userpatches variable below must be set to true for the above to work.
_user_patches="true"
# Apply all user patches without confirmation - !!! NOT RECOMMENDED !!!
_user_patches_no_confirm="false"

58
linux58-rc-tkg/linux58-tkg-config/ryzen-desktop-profile.cfg Normal file
View File

@ -0,0 +1,58 @@
# linux58-TkG config file
# Ryzen Desktop
#### MISC OPTIONS ####
# External config file to use - If the given file exists in path, it will override default config (customization.cfg) - Default is ~/.config/frogminer/linux52-tkg.cfg
_EXT_CONFIG_PATH=~/.config/frogminer/linux58-tkg.cfg
#### KERNEL OPTIONS ####
# Name of the default config file to use from the linux???-tkg-config folder. Arch default is "config.x86_64".
_configfile="config.x86_64"
# Disable some non-module debugging - See PKGBUILD for the list
_debugdisable="false"
# LEAVE AN EMPTY VALUE TO BE PROMPTED ABOUT FOLLOWING OPTIONS AT BUILD TIME
# Set to "true" to disable FUNCTION_TRACER/GRAPH_TRACER, lowering overhead but limiting debugging and analyzing of kernel functions - Kernel default is "false"
_ftracedisable="false"
# Set to "true" to disable NUMA, lowering overhead, but breaking CUDA/NvEnc on Nvidia equipped systems - Kernel default is "false"
_numadisable="false"
# Set to "true" to use explicit preemption points to lower latency at the cost of a small throughput loss - Can give a nice perf boost in VMs - Kernel default is "false"
_voluntary_preempt="false"
# A selection of patches from Zen/Liquorix kernel and additional tweaks for a better gaming experience (ZENIFY) - Default is "true"
_zenify="true"
# compiler optimization level - 1. Optimize for performance (-O2); 2. Optimize harder (-O3); 3. Optimize for size (-Os) - Kernel default is "2"
_compileroptlevel="1"
# Trust the CPU manufacturer to initialize Linux's CRNG (RANDOM_TRUST_CPU) - Kernel default is "false"
_random_trust_cpu="false"
# CPU scheduler runqueue sharing - No sharing (RQ_NONE), SMT (hyperthread) siblings (RQ_SMT), Multicore siblings (RQ_MC), Symmetric Multi-Processing (RQ_SMP), NUMA (RQ_ALL)
# Valid values are "none", "smt", "mc", "mc-llc"(for zen), "smp", "all" - Kernel default is "mc"
_runqueue_sharing="mc-llc"
# Timer frequency - "500", "750" or "1000" - More options available in kernel config prompt when left empty depending on selected cpusched - Kernel default is "500"
_timer_freq="500"
# Default CPU governor - "performance", "ondemand" (tweaked), "schedutil" or leave empty for default (schedutil on AMD and legacy Intel, intel_pstate on modern Intel) - Enforcing an option will disable intel_pstate altogether!
_default_cpu_gov="performance"
#### USER PATCHES ####
# You can use your own patches by putting them in the same folder as the PKGBUILD and giving them the .mypatch extension.
# You can also revert patches by putting them in the same folder as the PKGBUILD and giving them the .myrevert extension.
# Also, userpatches variable below must be set to true for the above to work.
_user_patches="true"
# Apply all user patches without confirmation - !!! NOT RECOMMENDED !!!
_user_patches_no_confirm="false"

156
linux58-rc-tkg/linux58-tkg-patches/0001-add-sysctl-to-disallow-unprivileged-CLONE_NEWUSER-by.patch Normal file
View File

@ -0,0 +1,156 @@
From 5ec2dd3a095442ec1a21d86042a4994f2ba24e63 Mon Sep 17 00:00:00 2001
Message-Id: <5ec2dd3a095442ec1a21d86042a4994f2ba24e63.1512651251.git.jan.steffens@gmail.com>
From: Serge Hallyn <serge.hallyn@canonical.com>
Date: Fri, 31 May 2013 19:12:12 +0100
Subject: [PATCH] add sysctl to disallow unprivileged CLONE_NEWUSER by default
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
[bwh: Remove unneeded binary sysctl bits]
Signed-off-by: Daniel Micay <danielmicay@gmail.com>
---
kernel/fork.c | 15 +++++++++++++++
kernel/sysctl.c | 12 ++++++++++++
kernel/user_namespace.c | 3 +++
3 files changed, 30 insertions(+)
diff --git a/kernel/fork.c b/kernel/fork.c
index 07cc743698d3668e..4011d68a8ff9305c 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -102,6 +102,11 @@
#define CREATE_TRACE_POINTS
#include <trace/events/task.h>
+#ifdef CONFIG_USER_NS
+extern int unprivileged_userns_clone;
+#else
+#define unprivileged_userns_clone 0
+#endif
/*
* Minimum number of threads to boot the kernel
@@ -1555,6 +1560,10 @@ static __latent_entropy struct task_struct *copy_process(
if ((clone_flags & (CLONE_NEWUSER|CLONE_FS)) == (CLONE_NEWUSER|CLONE_FS))
return ERR_PTR(-EINVAL);
+ if ((clone_flags & CLONE_NEWUSER) && !unprivileged_userns_clone)
+ if (!capable(CAP_SYS_ADMIN))
+ return ERR_PTR(-EPERM);
+
/*
* Thread groups must share signals as well, and detached threads
* can only be started up within the thread group.
@@ -2348,6 +2357,12 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags)
if (unshare_flags & CLONE_NEWNS)
unshare_flags |= CLONE_FS;
+ if ((unshare_flags & CLONE_NEWUSER) && !unprivileged_userns_clone) {
+ err = -EPERM;
+ if (!capable(CAP_SYS_ADMIN))
+ goto bad_unshare_out;
+ }
+
err = check_unshare_flags(unshare_flags);
if (err)
goto bad_unshare_out;
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index b86520ed3fb60fbf..f7dab3760839f1a1 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -105,6 +105,9 @@ extern int core_uses_pid;
#if defined(CONFIG_SYSCTL)
+#ifdef CONFIG_USER_NS
+extern int unprivileged_userns_clone;
+#endif
/* Constants used for minimum and maximum */
#ifdef CONFIG_LOCKUP_DETECTOR
static int sixty = 60;
@@ -513,6 +516,15 @@ static struct ctl_table kern_table[] = {
.proc_handler = proc_dointvec,
},
#endif
+#ifdef CONFIG_USER_NS
+ {
+ .procname = "unprivileged_userns_clone",
+ .data = &unprivileged_userns_clone,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ },
+#endif
#ifdef CONFIG_PROC_SYSCTL
{
.procname = "tainted",
diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c
index c490f1e4313b998a..dd03bd39d7bf194d 100644
--- a/kernel/user_namespace.c
+++ b/kernel/user_namespace.c
@@ -24,6 +24,9 @@
#include <linux/projid.h>
#include <linux/fs_struct.h>
+/* sysctl */
+int unprivileged_userns_clone;
+
static struct kmem_cache *user_ns_cachep __read_mostly;
static DEFINE_MUTEX(userns_state_mutex);
--
2.15.1
From b5202296055dd333db4425120d3f93ef4e6a0573 Mon Sep 17 00:00:00 2001
From: "Jan Alexander Steffens (heftig)" <jan.steffens@gmail.com>
Date: Thu, 7 Dec 2017 13:50:48 +0100
Subject: ZEN: Add CONFIG for unprivileged_userns_clone
This way our default behavior continues to match the vanilla kernel.
---
init/Kconfig | 16 ++++++++++++++++
kernel/user_namespace.c | 4 ++++
2 files changed, 20 insertions(+)
diff --git a/init/Kconfig b/init/Kconfig
index 4592bf7997c0..f3df02990aff 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -1004,6 +1004,22 @@ config USER_NS
If unsure, say N.
+config USER_NS_UNPRIVILEGED
+ bool "Allow unprivileged users to create namespaces"
+ default y
+ depends on USER_NS
+ help
+ When disabled, unprivileged users will not be able to create
+ new namespaces. Allowing users to create their own namespaces
+ has been part of several recent local privilege escalation
+ exploits, so if you need user namespaces but are
+ paranoid^Wsecurity-conscious you want to disable this.
+
+ This setting can be overridden at runtime via the
+ kernel.unprivileged_userns_clone sysctl.
+
+ If unsure, say Y.
+
config PID_NS
bool "PID Namespaces"
default y
diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c
index 6b9dbc257e34..107b17f0d528 100644
--- a/kernel/user_namespace.c
+++ b/kernel/user_namespace.c
@@ -27,7 +27,11 @@
#include <linux/sort.h>
/* sysctl */
+#ifdef CONFIG_USER_NS_UNPRIVILEGED
+int unprivileged_userns_clone = 1;
+#else
int unprivileged_userns_clone;
+#endif
static struct kmem_cache *user_ns_cachep __read_mostly;
static DEFINE_MUTEX(userns_state_mutex);

354
linux58-rc-tkg/linux58-tkg-patches/0002-clear-patches.patch Normal file
View File

@ -0,0 +1,354 @@
From 2ac70785613ef4c6b16414986bb18bd7b60d2a13 Mon Sep 17 00:00:00 2001
From: Arjan van de Ven <arjan@linux.intel.com>
Date: Mon, 14 Mar 2016 11:10:58 -0600
Subject: [PATCH] pci pme wakeups
Reduce wakeups for PME checks, which are a workaround for miswired
boards (sadly, too many of them) in laptops.
---
drivers/pci/pci.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index c25acace7d91..0ddebdad9f5b 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -61,7 +61,7 @@ struct pci_pme_device {
struct pci_dev *dev;
};
-#define PME_TIMEOUT 1000 /* How long between PME checks */
+#define PME_TIMEOUT 4000 /* How long between PME checks */
static void pci_dev_d3_sleep(struct pci_dev *dev)
{
--
2.20.1
From 7e7e36c67aa71d6a1ec5676d99d37c1fea389ceb Mon Sep 17 00:00:00 2001
From: Arjan van de Ven <arjan@linux.intel.com>
Date: Sat, 19 Mar 2016 21:32:19 -0400
Subject: [PATCH] intel_idle: tweak cpuidle cstates
Increase target_residency in cpuidle cstate
Tune intel_idle to be a bit less agressive;
Clear linux is cleaner in hygiene (wakupes) than the average linux,
so we can afford changing these in a way that increases
performance while keeping power efficiency
---
drivers/idle/intel_idle.c | 44 +++++++++++++++++++--------------------
1 file changed, 22 insertions(+), 22 deletions(-)
diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
index 8b5d85c91e9d..5e2d813a048d 100644
--- a/drivers/idle/intel_idle.c
+++ b/drivers/idle/intel_idle.c
@@ -466,7 +466,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x01",
.flags = MWAIT2flg(0x01),
.exit_latency = 10,
- .target_residency = 20,
+ .target_residency = 120,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -474,7 +474,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x10",
.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 33,
- .target_residency = 100,
+ .target_residency = 900,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -482,7 +482,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x20",
.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 133,
- .target_residency = 400,
+ .target_residency = 1000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -490,7 +490,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x32",
.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 166,
- .target_residency = 500,
+ .target_residency = 1500,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -498,7 +498,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x40",
.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 300,
- .target_residency = 900,
+ .target_residency = 2000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -506,7 +506,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x50",
.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 600,
- .target_residency = 1800,
+ .target_residency = 5000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -514,7 +514,7 @@ static struct cpuidle_state hsw_cstates[] = {
.desc = "MWAIT 0x60",
.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 2600,
- .target_residency = 7700,
+ .target_residency = 9000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -534,7 +534,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x01",
.flags = MWAIT2flg(0x01),
.exit_latency = 10,
- .target_residency = 20,
+ .target_residency = 120,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -542,7 +542,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x10",
.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 40,
- .target_residency = 100,
+ .target_residency = 1000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -550,7 +550,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x20",
.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 133,
- .target_residency = 400,
+ .target_residency = 1000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -558,7 +558,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x32",
.flags = MWAIT2flg(0x32) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 166,
- .target_residency = 500,
+ .target_residency = 2000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -566,7 +566,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x40",
.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 300,
- .target_residency = 900,
+ .target_residency = 4000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -574,7 +574,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x50",
.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 600,
- .target_residency = 1800,
+ .target_residency = 7000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -582,7 +582,7 @@ static struct cpuidle_state bdw_cstates[] = {
.desc = "MWAIT 0x60",
.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 2600,
- .target_residency = 7700,
+ .target_residency = 9000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -603,7 +603,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x01",
.flags = MWAIT2flg(0x01),
.exit_latency = 10,
- .target_residency = 20,
+ .target_residency = 120,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -611,7 +611,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x10",
.flags = MWAIT2flg(0x10) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 70,
- .target_residency = 100,
+ .target_residency = 1000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -619,7 +619,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x20",
.flags = MWAIT2flg(0x20) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 85,
- .target_residency = 200,
+ .target_residency = 600,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -627,7 +627,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x33",
.flags = MWAIT2flg(0x33) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 124,
- .target_residency = 800,
+ .target_residency = 3000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -635,7 +635,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x40",
.flags = MWAIT2flg(0x40) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 200,
- .target_residency = 800,
+ .target_residency = 3200,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -643,7 +643,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x50",
.flags = MWAIT2flg(0x50) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 480,
- .target_residency = 5000,
+ .target_residency = 9000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -651,7 +651,7 @@ static struct cpuidle_state skl_cstates[] = {
.desc = "MWAIT 0x60",
.flags = MWAIT2flg(0x60) | CPUIDLE_FLAG_TLB_FLUSHED,
.exit_latency = 890,
- .target_residency = 5000,
+ .target_residency = 9000,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
@@ -672,7 +672,7 @@ static struct cpuidle_state skx_cstates[] = {
.desc = "MWAIT 0x01",
.flags = MWAIT2flg(0x01),
.exit_latency = 10,
- .target_residency = 20,
+ .target_residency = 300,
.enter = &intel_idle,
.enter_s2idle = intel_idle_s2idle, },
{
--
2.20.1
From b8211d4f79dd88dfc2d4bd52be46103ea0b70e3e Mon Sep 17 00:00:00 2001
From: Arjan van de Ven <arjan@linux.intel.com>
Date: Fri, 6 Jan 2017 15:34:09 +0000
Subject: [PATCH] ipv4/tcp: allow the memory tuning for tcp to go a little
bigger than default
---
net/ipv4/tcp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index cf3c5095c10e..b30d51837b2d 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -3897,8 +3897,8 @@ void __init tcp_init(void)
tcp_init_mem();
/* Set per-socket limits to no more than 1/128 the pressure threshold */
limit = nr_free_buffer_pages() << (PAGE_SHIFT - 7);
- max_wshare = min(4UL*1024*1024, limit);
- max_rshare = min(6UL*1024*1024, limit);
+ max_wshare = min(16UL*1024*1024, limit);
+ max_rshare = min(16UL*1024*1024, limit);
init_net.ipv4.sysctl_tcp_wmem[0] = SK_MEM_QUANTUM;
init_net.ipv4.sysctl_tcp_wmem[1] = 16*1024;
--
2.20.1
From 050223869257b87e22636158a80da38d877248ed Mon Sep 17 00:00:00 2001
From: Arjan van de Ven <arjan@linux.intel.com>
Date: Sun, 18 Feb 2018 23:35:41 +0000
Subject: [PATCH] locking: rwsem: spin faster
tweak rwsem owner spinning a bit
---
kernel/locking/rwsem.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/kernel/locking/rwsem.c b/kernel/locking/rwsem.c
index eef04551eae7..1ec5ab4c8ff7 100644
--- a/kernel/locking/rwsem.c
+++ b/kernel/locking/rwsem.c
@@ -720,6 +720,7 @@ rwsem_spin_on_owner(struct rw_semaphore *sem, unsigned long nonspinnable)
struct task_struct *new, *owner;
unsigned long flags, new_flags;
enum owner_state state;
+ int i = 0;
owner = rwsem_owner_flags(sem, &flags);
state = rwsem_owner_state(owner, flags, nonspinnable);
@@ -753,7 +754,8 @@ rwsem_spin_on_owner(struct rw_semaphore *sem, unsigned long nonspinnable)
break;
}
- cpu_relax();
+ if (i++ > 1000)
+ cpu_relax();
}
rcu_read_unlock();
From b836ea320114643d4354b43acb6ec8bb06ada487 Mon Sep 17 00:00:00 2001
From: Arjan van de Ven <arjan@linux.intel.com>
Date: Thu, 2 Jun 2016 23:36:32 -0500
Subject: [PATCH] drivers: Initialize ata before graphics
ATA init is the long pole in the boot process, and its asynchronous.
move the graphics init after it so that ata and graphics initialize
in parallel
---
drivers/Makefile | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/Makefile b/drivers/Makefile
index aaef17cc6512..d08f3a394929 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -58,15 +58,8 @@ obj-y += char/
# iommu/ comes before gpu as gpu are using iommu controllers
obj-y += iommu/
-# gpu/ comes after char for AGP vs DRM startup and after iommu
-obj-y += gpu/
-
obj-$(CONFIG_CONNECTOR) += connector/
-# i810fb and intelfb depend on char/agp/
-obj-$(CONFIG_FB_I810) += video/fbdev/i810/
-obj-$(CONFIG_FB_INTEL) += video/fbdev/intelfb/
-
obj-$(CONFIG_PARPORT) += parport/
obj-$(CONFIG_NVM) += lightnvm/
obj-y += base/ block/ misc/ mfd/ nfc/
@@ -79,6 +72,14 @@ obj-$(CONFIG_IDE) += ide/
obj-y += scsi/
obj-y += nvme/
obj-$(CONFIG_ATA) += ata/
+
+# gpu/ comes after char for AGP vs DRM startup and after iommu
+obj-y += gpu/
+
+# i810fb and intelfb depend on char/agp/
+obj-$(CONFIG_FB_I810) += video/fbdev/i810/
+obj-$(CONFIG_FB_INTEL) += video/fbdev/intelfb/
+
obj-$(CONFIG_TARGET_CORE) += target/
obj-$(CONFIG_MTD) += mtd/
obj-$(CONFIG_SPI) += spi/

1446
linux58-rc-tkg/linux58-tkg-patches/0003-glitched-base.patch Normal file
View File

File diff suppressed because it is too large Load Diff

72
linux58-rc-tkg/linux58-tkg-patches/0003-glitched-cfs.patch Normal file
View File

@ -0,0 +1,72 @@
diff --git a/kernel/Kconfig.hz b/kernel/Kconfig.hz
index 2a202a846757..1d9c7ed79b11 100644
--- a/kernel/Kconfig.hz
+++ b/kernel/Kconfig.hz
@@ -4,7 +4,7 @@
choice
prompt "Timer frequency"
- default HZ_250
+ default HZ_500
help
Allows the configuration of the timer frequency. It is customary
to have the timer interrupt run at 1000 Hz but 100 Hz may be more
@@ -39,6 +39,13 @@ choice
on SMP and NUMA systems and exactly dividing by both PAL and
NTSC frame rates for video and multimedia work.
+ config HZ_500
+ bool "500 HZ"
+ help
+ 500 Hz is a balanced timer frequency. Provides fast interactivity
+ on desktops with great smoothness without increasing CPU power
+ consumption and sacrificing the battery life on laptops.
+
config HZ_1000
bool "1000 HZ"
help
@@ -52,6 +59,7 @@ config HZ
default 100 if HZ_100
default 250 if HZ_250
default 300 if HZ_300
+ default 500 if HZ_500
default 1000 if HZ_1000
config SCHED_HRTICK
diff --git a/kernel/Kconfig.hz b/kernel/Kconfig.hz
index 2a202a846757..1d9c7ed79b11 100644
--- a/kernel/Kconfig.hz
+++ b/kernel/Kconfig.hz
@@ -4,7 +4,7 @@
choice
prompt "Timer frequency"
- default HZ_500
+ default HZ_750
help
Allows the configuration of the timer frequency. It is customary
to have the timer interrupt run at 1000 Hz but 100 Hz may be more
@@ -46,6 +46,13 @@ choice
on desktops with great smoothness without increasing CPU power
consumption and sacrificing the battery life on laptops.
+ config HZ_750
+ bool "750 HZ"
+ help
+ 750 Hz is a good timer frequency for desktops. Provides fast
+ interactivity with great smoothness without sacrificing too
+ much throughput.
+
config HZ_1000
bool "1000 HZ"
help
@@ -60,6 +67,7 @@ config HZ
default 250 if HZ_250
default 300 if HZ_300
default 500 if HZ_500
+ default 750 if HZ_750
default 1000 if HZ_1000
config SCHED_HRTICK

18
linux58-rc-tkg/linux58-tkg-patches/0005-glitched-ondemand-pds.patch Normal file
View File

@ -0,0 +1,18 @@
diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 6b423eebfd5d..61e3271675d6 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -21,10 +21,10 @@
#include "cpufreq_ondemand.h"
/* On-demand governor macros */
-#define DEF_FREQUENCY_UP_THRESHOLD (63)
-#define DEF_SAMPLING_DOWN_FACTOR (1)
+#define DEF_FREQUENCY_UP_THRESHOLD (55)
+#define DEF_SAMPLING_DOWN_FACTOR (5)
#define MAX_SAMPLING_DOWN_FACTOR (100000)
-#define MICRO_FREQUENCY_UP_THRESHOLD (95)
+#define MICRO_FREQUENCY_UP_THRESHOLD (63)
#define MICRO_FREQUENCY_MIN_SAMPLE_RATE (10000)
#define MIN_FREQUENCY_UP_THRESHOLD (1)
#define MAX_FREQUENCY_UP_THRESHOLD (100)

166
linux58-rc-tkg/linux58-tkg-patches/0005-glitched-pds.patch Normal file
View File

@ -0,0 +1,166 @@
From f7f49141a5dbe9c99d78196b58c44307fb2e6be3 Mon Sep 17 00:00:00 2001
From: Tk-Glitch <ti3nou@gmail.com>
Date: Wed, 4 Jul 2018 04:30:08 +0200
Subject: glitched - PDS
diff --git a/kernel/Kconfig.hz b/kernel/Kconfig.hz
index 2a202a846757..1d9c7ed79b11 100644
--- a/kernel/Kconfig.hz
+++ b/kernel/Kconfig.hz
@@ -4,7 +4,7 @@
choice
prompt "Timer frequency"
- default HZ_250
+ default HZ_500
help
Allows the configuration of the timer frequency. It is customary
to have the timer interrupt run at 1000 Hz but 100 Hz may be more
@@ -39,6 +39,13 @@ choice
on SMP and NUMA systems and exactly dividing by both PAL and
NTSC frame rates for video and multimedia work.
+ config HZ_500
+ bool "500 HZ"
+ help
+ 500 Hz is a balanced timer frequency. Provides fast interactivity
+ on desktops with great smoothness without increasing CPU power
+ consumption and sacrificing the battery life on laptops.
+
config HZ_1000
bool "1000 HZ"
help
@@ -52,6 +59,7 @@ config HZ
default 100 if HZ_100
default 250 if HZ_250
default 300 if HZ_300
+ default 500 if HZ_500
default 1000 if HZ_1000
config SCHED_HRTICK
diff --git a/kernel/Kconfig.hz b/kernel/Kconfig.hz
index 2a202a846757..1d9c7ed79b11 100644
--- a/kernel/Kconfig.hz
+++ b/kernel/Kconfig.hz
@@ -4,7 +4,7 @@
choice
prompt "Timer frequency"
- default HZ_500
+ default HZ_750
help
Allows the configuration of the timer frequency. It is customary
to have the timer interrupt run at 1000 Hz but 100 Hz may be more
@@ -46,6 +46,13 @@ choice
on desktops with great smoothness without increasing CPU power
consumption and sacrificing the battery life on laptops.
+ config HZ_750
+ bool "750 HZ"
+ help
+ 750 Hz is a good timer frequency for desktops. Provides fast
+ interactivity with great smoothness without sacrificing too
+ much throughput.
+
config HZ_1000
bool "1000 HZ"
help
@@ -60,6 +67,7 @@ config HZ
default 250 if HZ_250
default 300 if HZ_300
default 500 if HZ_500
+ default 750 if HZ_750
default 1000 if HZ_1000
config SCHED_HRTICK
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 9270a4370d54..30d01e647417 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -159,7 +159,7 @@ struct scan_control {
/*
* From 0 .. 100. Higher means more swappy.
*/
-int vm_swappiness = 60;
+int vm_swappiness = 20;
/*
* The total number of pages which are beyond the high watermark within all
* zones.
diff --git a/init/Kconfig b/init/Kconfig
index 11fd9b502d06..e9bc34d3019b 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -715,6 +715,7 @@ menu "Scheduler features"
config UCLAMP_TASK
bool "Enable utilization clamping for RT/FAIR tasks"
depends on CPU_FREQ_GOV_SCHEDUTIL
+ depends on !SCHED_PDS
help
This feature enables the scheduler to track the clamped utilization
of each CPU based on RUNNABLE tasks scheduled on that CPU.
@@ -948,7 +948,6 @@ config CGROUP_DEVICE
config CGROUP_CPUACCT
bool "Simple CPU accounting controller"
- depends on !SCHED_PDS
help
Provides a simple controller for monitoring the
total CPU consumed by the tasks in a cgroup.
diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile
index b23231bae996..cab4e5c5b38e 100644
--- a/kernel/sched/Makefile
+++ b/kernel/sched/Makefile
@@ -24,13 +24,13 @@ obj-y += fair.o rt.o deadline.o
obj-$(CONFIG_SMP) += cpudeadline.o topology.o stop_task.o
obj-$(CONFIG_SCHED_AUTOGROUP) += autogroup.o
obj-$(CONFIG_SCHED_DEBUG) += debug.o
-obj-$(CONFIG_CGROUP_CPUACCT) += cpuacct.o
endif
obj-y += loadavg.o clock.o cputime.o
obj-y += idle.o
obj-y += wait.o wait_bit.o swait.o completion.o
obj-$(CONFIG_SMP) += cpupri.o pelt.o
obj-$(CONFIG_SCHEDSTATS) += stats.o
+obj-$(CONFIG_CGROUP_CPUACCT) += cpuacct.o
obj-$(CONFIG_CPU_FREQ) += cpufreq.o
obj-$(CONFIG_CPU_FREQ_GOV_SCHEDUTIL) += cpufreq_schedutil.o
obj-$(CONFIG_MEMBARRIER) += membarrier.o
diff --git a/kernel/sched/pds.c b/kernel/sched/pds.c
index 9281ad164..f09a609cf 100644
--- a/kernel/sched/pds.c
+++ b/kernel/sched/pds.c
@@ -81,6 +81,18 @@ enum {
NR_CPU_AFFINITY_CHK_LEVEL
};
+/*
+ * This allows printing both to /proc/sched_debug and
+ * to the console
+ */
+#define SEQ_printf(m, x...) \
+ do { \
+ if (m) \
+ seq_printf(m, x); \
+ else \
+ pr_cont(x); \
+ } while (0)
+
static inline void print_scheduler_version(void)
{
printk(KERN_INFO "pds: PDS-mq CPU Scheduler 0.99o by Alfred Chen.\n");
@@ -6353,7 +6365,10 @@ void ia64_set_curr_task(int cpu, struct task_struct *p)
#ifdef CONFIG_SCHED_DEBUG
void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns,
struct seq_file *m)
-{}
+{
+ SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, task_pid_nr_ns(p, ns),
+ get_nr_threads(p));
+}
void proc_sched_set_task(struct task_struct *p)
{}

8542
linux58-rc-tkg/linux58-tkg-patches/0005-v5.8_undead-pds099o.patch Normal file
View File

File diff suppressed because it is too large Load Diff

193
linux58-rc-tkg/linux58-tkg-patches/0006-add-acs-overrides_iommu.patch Normal file
View File

@ -0,0 +1,193 @@
From cdeab384f48dd9c88e2dff2e9ad8d57dca1a1b1c Mon Sep 17 00:00:00 2001
From: Mark Weiman <mark.weiman@markzz.com>
Date: Sun, 12 Aug 2018 11:36:21 -0400
Subject: [PATCH] pci: Enable overrides for missing ACS capabilities
This an updated version of Alex Williamson's patch from:
https://lkml.org/lkml/2013/5/30/513
Original commit message follows:
PCIe ACS (Access Control Services) is the PCIe 2.0+ feature that
allows us to control whether transactions are allowed to be redirected
in various subnodes of a PCIe topology. For instance, if two
endpoints are below a root port or downsteam switch port, the
downstream port may optionally redirect transactions between the
devices, bypassing upstream devices. The same can happen internally
on multifunction devices. The transaction may never be visible to the
upstream devices.
One upstream device that we particularly care about is the IOMMU. If
a redirection occurs in the topology below the IOMMU, then the IOMMU
cannot provide isolation between devices. This is why the PCIe spec
encourages topologies to include ACS support. Without it, we have to
assume peer-to-peer DMA within a hierarchy can bypass IOMMU isolation.
Unfortunately, far too many topologies do not support ACS to make this
a steadfast requirement. Even the latest chipsets from Intel are only
sporadically supporting ACS. We have trouble getting interconnect
vendors to include the PCIe spec required PCIe capability, let alone
suggested features.
Therefore, we need to add some flexibility. The pcie_acs_override=
boot option lets users opt-in specific devices or sets of devices to
assume ACS support. The "downstream" option assumes full ACS support
on root ports and downstream switch ports. The "multifunction"
option assumes the subset of ACS features available on multifunction
endpoints and upstream switch ports are supported. The "id:nnnn:nnnn"
option enables ACS support on devices matching the provided vendor
and device IDs, allowing more strategic ACS overrides. These options
may be combined in any order. A maximum of 16 id specific overrides
are available. It's suggested to use the most limited set of options
necessary to avoid completely disabling ACS across the topology.
Note to hardware vendors, we have facilities to permanently quirk
specific devices which enforce isolation but not provide an ACS
capability. Please contact me to have your devices added and save
your customers the hassle of this boot option.
Signed-off-by: Mark Weiman <mark.weiman@markzz.com>
---
.../admin-guide/kernel-parameters.txt | 9 ++
drivers/pci/quirks.c | 101 ++++++++++++++++++
2 files changed, 110 insertions(+)
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index aefd358a5ca3..173b3596fd9e 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -3190,6 +3190,15 @@
nomsi [MSI] If the PCI_MSI kernel config parameter is
enabled, this kernel boot option can be used to
disable the use of MSI interrupts system-wide.
+ pcie_acs_override =
+ [PCIE] Override missing PCIe ACS support for:
+ downstream
+ All downstream ports - full ACS capabilities
+ multifunction
+ All multifunction devices - multifunction ACS subset
+ id:nnnn:nnnn
+ Specific device - full ACS capabilities
+ Specified as vid:did (vendor/device ID) in hex
noioapicquirk [APIC] Disable all boot interrupt quirks.
Safety option to keep boot IRQs enabled. This
should never be necessary.
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 4700d24e5d55..8f7a3d7fd9c1 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -3372,6 +3372,106 @@ static void quirk_no_bus_reset(struct pci_dev *dev)
dev->dev_flags |= PCI_DEV_FLAGS_NO_BUS_RESET;
}
+static bool acs_on_downstream;
+static bool acs_on_multifunction;
+
+#define NUM_ACS_IDS 16
+struct acs_on_id {
+ unsigned short vendor;
+ unsigned short device;
+};
+static struct acs_on_id acs_on_ids[NUM_ACS_IDS];
+static u8 max_acs_id;
+
+static __init int pcie_acs_override_setup(char *p)
+{
+ if (!p)
+ return -EINVAL;
+
+ while (*p) {
+ if (!strncmp(p, "downstream", 10))
+ acs_on_downstream = true;
+ if (!strncmp(p, "multifunction", 13))
+ acs_on_multifunction = true;
+ if (!strncmp(p, "id:", 3)) {
+ char opt[5];
+ int ret;
+ long val;
+
+ if (max_acs_id >= NUM_ACS_IDS - 1) {
+ pr_warn("Out of PCIe ACS override slots (%d)\n",
+ NUM_ACS_IDS);
+ goto next;
+ }
+
+ p += 3;
+ snprintf(opt, 5, "%s", p);
+ ret = kstrtol(opt, 16, &val);
+ if (ret) {
+ pr_warn("PCIe ACS ID parse error %d\n", ret);
+ goto next;
+ }
+ acs_on_ids[max_acs_id].vendor = val;
+
+ p += strcspn(p, ":");
+ if (*p != ':') {
+ pr_warn("PCIe ACS invalid ID\n");
+ goto next;
+ }
+
+ p++;
+ snprintf(opt, 5, "%s", p);
+ ret = kstrtol(opt, 16, &val);
+ if (ret) {
+ pr_warn("PCIe ACS ID parse error %d\n", ret);
+ goto next;
+ }
+ acs_on_ids[max_acs_id].device = val;
+ max_acs_id++;
+ }
+next:
+ p += strcspn(p, ",");
+ if (*p == ',')
+ p++;
+ }
+
+ if (acs_on_downstream || acs_on_multifunction || max_acs_id)
+ pr_warn("Warning: PCIe ACS overrides enabled; This may allow non-IOMMU protected peer-to-peer DMA\n");
+
+ return 0;
+}
+early_param("pcie_acs_override", pcie_acs_override_setup);
+
+static int pcie_acs_overrides(struct pci_dev *dev, u16 acs_flags)
+{
+ int i;
+
+ /* Never override ACS for legacy devices or devices with ACS caps */
+ if (!pci_is_pcie(dev) ||
+ pci_find_ext_capability(dev, PCI_EXT_CAP_ID_ACS))
+ return -ENOTTY;
+
+ for (i = 0; i < max_acs_id; i++)
+ if (acs_on_ids[i].vendor == dev->vendor &&
+ acs_on_ids[i].device == dev->device)
+ return 1;
+
+ switch (pci_pcie_type(dev)) {
+ case PCI_EXP_TYPE_DOWNSTREAM:
+ case PCI_EXP_TYPE_ROOT_PORT:
+ if (acs_on_downstream)
+ return 1;
+ break;
+ case PCI_EXP_TYPE_ENDPOINT:
+ case PCI_EXP_TYPE_UPSTREAM:
+ case PCI_EXP_TYPE_LEG_END:
+ case PCI_EXP_TYPE_RC_END:
+ if (acs_on_multifunction && dev->multifunction)
+ return 1;
+ }
+
+ return -ENOTTY;
+}
/*
* Some Atheros AR9xxx and QCA988x chips do not behave after a bus reset.
* The device will throw a Link Down error on AER-capable systems and
@@ -4513,6 +4613,7 @@ static const struct pci_dev_acs_enabled {
{ PCI_VENDOR_ID_ZHAOXIN, 0x9083, pci_quirk_mf_endpoint_acs },
/* Zhaoxin Root/Downstream Ports */
{ PCI_VENDOR_ID_ZHAOXIN, PCI_ANY_ID, pci_quirk_zhaoxin_pcie_ports_acs },
+ { PCI_ANY_ID, PCI_ANY_ID, pcie_acs_overrides },
{ 0 }
};

908
linux58-rc-tkg/linux58-tkg-patches/0007-v5.8-fsync.patch Normal file
View File

@ -0,0 +1,908 @@
From f7f49141a5dbe9c99d78196b58c44307fb2e6be3 Mon Sep 17 00:00:00 2001
From: Tk-Glitch <ti3nou@gmail.com>
Date: Mon, 20 Apr 2020 14:09:11 +0200
Subject: Import Fsync v3 patchset - Squashed from https://gitlab.collabora.com/tonyk/linux/-/commits/futex-proton-v3
diff --git a/include/uapi/linux/futex.h b/include/uapi/linux/futex.h
index a89eb0accd5e2ee527be1e3e11b1117ff5bf94b4..580001e89c6caed57dd8b3cb491d65dce846caff 100644
--- a/include/uapi/linux/futex.h
+++ b/include/uapi/linux/futex.h
@@ -21,6 +21,7 @@
#define FUTEX_WAKE_BITSET 10
#define FUTEX_WAIT_REQUEUE_PI 11
#define FUTEX_CMP_REQUEUE_PI 12
+#define FUTEX_WAIT_MULTIPLE 13
#define FUTEX_PRIVATE_FLAG 128
#define FUTEX_CLOCK_REALTIME 256
@@ -40,6 +41,8 @@
FUTEX_PRIVATE_FLAG)
#define FUTEX_CMP_REQUEUE_PI_PRIVATE (FUTEX_CMP_REQUEUE_PI | \
FUTEX_PRIVATE_FLAG)
+#define FUTEX_WAIT_MULTIPLE_PRIVATE (FUTEX_WAIT_MULTIPLE | \
+ FUTEX_PRIVATE_FLAG)
/*
* Support for robust futexes: the kernel cleans up held futexes at
@@ -150,4 +153,21 @@ struct robust_list_head {
(((op & 0xf) << 28) | ((cmp & 0xf) << 24) \
| ((oparg & 0xfff) << 12) | (cmparg & 0xfff))
+/*
+ * Maximum number of multiple futexes to wait for
+ */
+#define FUTEX_MULTIPLE_MAX_COUNT 128
+
+/**
+ * struct futex_wait_block - Block of futexes to be waited for
+ * @uaddr: User address of the futex
+ * @val: Futex value expected by userspace
+ * @bitset: Bitset for the optional bitmasked wakeup
+ */
+struct futex_wait_block {
+ __u32 __user *uaddr;
+ __u32 val;
+ __u32 bitset;
+};
+
#endif /* _UAPI_LINUX_FUTEX_H */
diff --git a/kernel/futex.c b/kernel/futex.c
index 0cf84c8664f207c574325b899ef2e57f01295a94..58cf9eb2b851b4858e29b5ef4114a29a92e676ba 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -215,6 +215,8 @@ struct futex_pi_state {
* @rt_waiter: rt_waiter storage for use with requeue_pi
* @requeue_pi_key: the requeue_pi target futex key
* @bitset: bitset for the optional bitmasked wakeup
+ * @uaddr: userspace address of futex
+ * @uval: expected futex's value
*
* We use this hashed waitqueue, instead of a normal wait_queue_entry_t, so
* we can wake only the relevant ones (hashed queues may be shared).
@@ -237,6 +239,8 @@ struct futex_q {
struct rt_mutex_waiter *rt_waiter;
union futex_key *requeue_pi_key;
u32 bitset;
+ u32 __user *uaddr;
+ u32 uval;
} __randomize_layout;
static const struct futex_q futex_q_init = {
@@ -2420,6 +2424,29 @@ static int unqueue_me(struct futex_q *q)
return ret;
}
+/**
+ * unqueue_multiple() - Remove several futexes from their futex_hash_bucket
+ * @q: The list of futexes to unqueue
+ * @count: Number of futexes in the list
+ *
+ * Helper to unqueue a list of futexes. This can't fail.
+ *
+ * Return:
+ * - >=0 - Index of the last futex that was awoken;
+ * - -1 - If no futex was awoken
+ */
+static int unqueue_multiple(struct futex_q *q, int count)
+{
+ int ret = -1;
+ int i;
+
+ for (i = 0; i < count; i++) {
+ if (!unqueue_me(&q[i]))
+ ret = i;
+ }
+ return ret;
+}
+
/*
* PI futexes can not be requeued and must remove themself from the
* hash bucket. The hash bucket lock (i.e. lock_ptr) is held on entry
@@ -2783,6 +2810,211 @@ static int futex_wait_setup(u32 __user *uaddr, u32 val, unsigned int flags,
return ret;
}
+/**
+ * futex_wait_multiple_setup() - Prepare to wait and enqueue multiple futexes
+ * @qs: The corresponding futex list
+ * @count: The size of the lists
+ * @flags: Futex flags (FLAGS_SHARED, etc.)
+ * @awaken: Index of the last awoken futex
+ *
+ * Prepare multiple futexes in a single step and enqueue them. This may fail if
+ * the futex list is invalid or if any futex was already awoken. On success the
+ * task is ready to interruptible sleep.
+ *
+ * Return:
+ * - 1 - One of the futexes was awaken by another thread
+ * - 0 - Success
+ * - <0 - -EFAULT, -EWOULDBLOCK or -EINVAL
+ */
+static int futex_wait_multiple_setup(struct futex_q *qs, int count,
+ unsigned int flags, int *awaken)
+{
+ struct futex_hash_bucket *hb;
+ int ret, i;
+ u32 uval;
+
+ /*
+ * Enqueuing multiple futexes is tricky, because we need to
+ * enqueue each futex in the list before dealing with the next
+ * one to avoid deadlocking on the hash bucket. But, before
+ * enqueuing, we need to make sure that current->state is
+ * TASK_INTERRUPTIBLE, so we don't absorb any awake events, which
+ * cannot be done before the get_futex_key of the next key,
+ * because it calls get_user_pages, which can sleep. Thus, we
+ * fetch the list of futexes keys in two steps, by first pinning
+ * all the memory keys in the futex key, and only then we read
+ * each key and queue the corresponding futex.
+ */
+retry:
+ for (i = 0; i < count; i++) {
+ qs[i].key = FUTEX_KEY_INIT;
+ ret = get_futex_key(qs[i].uaddr, flags & FLAGS_SHARED,
+ &qs[i].key, FUTEX_READ);
+ if (unlikely(ret)) {
+ for (--i; i >= 0; i--)
+ put_futex_key(&qs[i].key);
+ return ret;
+ }
+ }
+
+ set_current_state(TASK_INTERRUPTIBLE);
+
+ for (i = 0; i < count; i++) {
+ struct futex_q *q = &qs[i];
+
+ hb = queue_lock(q);
+
+ ret = get_futex_value_locked(&uval, q->uaddr);
+ if (ret) {
+ /*
+ * We need to try to handle the fault, which
+ * cannot be done without sleep, so we need to
+ * undo all the work already done, to make sure
+ * we don't miss any wake ups. Therefore, clean
+ * up, handle the fault and retry from the
+ * beginning.
+ */
+ queue_unlock(hb);
+
+ /*
+ * Keys 0..(i-1) are implicitly put
+ * on unqueue_multiple.
+ */
+ put_futex_key(&q->key);
+
+ *awaken = unqueue_multiple(qs, i);
+
+ __set_current_state(TASK_RUNNING);
+
+ /*
+ * On a real fault, prioritize the error even if
+ * some other futex was awoken. Userspace gave
+ * us a bad address, -EFAULT them.
+ */
+ ret = get_user(uval, q->uaddr);
+ if (ret)
+ return ret;
+
+ /*
+ * Even if the page fault was handled, If
+ * something was already awaken, we can safely
+ * give up and succeed to give a hint for userspace to
+ * acquire the right futex faster.
+ */
+ if (*awaken >= 0)
+ return 1;
+
+ goto retry;
+ }
+
+ if (uval != q->uval) {
+ queue_unlock(hb);
+
+ put_futex_key(&qs[i].key);
+
+ /*
+ * If something was already awaken, we can
+ * safely ignore the error and succeed.
+ */
+ *awaken = unqueue_multiple(qs, i);
+ __set_current_state(TASK_RUNNING);
+ if (*awaken >= 0)
+ return 1;
+
+ return -EWOULDBLOCK;
+ }
+
+ /*
+ * The bucket lock can't be held while dealing with the
+ * next futex. Queue each futex at this moment so hb can
+ * be unlocked.
+ */
+ queue_me(&qs[i], hb);
+ }
+ return 0;
+}
+
+/**
+ * futex_wait_multiple() - Prepare to wait on and enqueue several futexes
+ * @qs: The list of futexes to wait on
+ * @op: Operation code from futex's syscall
+ * @count: The number of objects
+ * @abs_time: Timeout before giving up and returning to userspace
+ *
+ * Entry point for the FUTEX_WAIT_MULTIPLE futex operation, this function
+ * sleeps on a group of futexes and returns on the first futex that
+ * triggered, or after the timeout has elapsed.
+ *
+ * Return:
+ * - >=0 - Hint to the futex that was awoken
+ * - <0 - On error
+ */
+static int futex_wait_multiple(struct futex_q *qs, int op,
+ u32 count, ktime_t *abs_time)
+{
+ struct hrtimer_sleeper timeout, *to;
+ int ret, flags = 0, hint = 0;
+ unsigned int i;
+
+ if (!(op & FUTEX_PRIVATE_FLAG))
+ flags |= FLAGS_SHARED;
+
+ if (op & FUTEX_CLOCK_REALTIME)
+ flags |= FLAGS_CLOCKRT;
+
+ to = futex_setup_timer(abs_time, &timeout, flags, 0);
+ while (1) {
+ ret = futex_wait_multiple_setup(qs, count, flags, &hint);
+ if (ret) {
+ if (ret > 0) {
+ /* A futex was awaken during setup */
+ ret = hint;
+ }
+ break;
+ }
+
+ if (to)
+ hrtimer_start_expires(&to->timer, HRTIMER_MODE_ABS);
+
+ /*
+ * Avoid sleeping if another thread already tried to
+ * wake us.
+ */
+ for (i = 0; i < count; i++) {
+ if (plist_node_empty(&qs[i].list))
+ break;
+ }
+
+ if (i == count && (!to || to->task))
+ freezable_schedule();
+
+ ret = unqueue_multiple(qs, count);
+
+ __set_current_state(TASK_RUNNING);
+
+ if (ret >= 0)
+ break;
+ if (to && !to->task) {
+ ret = -ETIMEDOUT;
+ break;
+ } else if (signal_pending(current)) {
+ ret = -ERESTARTSYS;
+ break;
+ }
+ /*
+ * The final case is a spurious wakeup, for
+ * which just retry.
+ */
+ }
+
+ if (to) {
+ hrtimer_cancel(&to->timer);
+ destroy_hrtimer_on_stack(&to->timer);
+ }
+
+ return ret;
+}
+
static int futex_wait(u32 __user *uaddr, unsigned int flags, u32 val,
ktime_t *abs_time, u32 bitset)
{
@@ -3907,6 +4139,43 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout,
return -ENOSYS;
}
+/**
+ * futex_read_wait_block - Read an array of futex_wait_block from userspace
+ * @uaddr: Userspace address of the block
+ * @count: Number of blocks to be read
+ *
+ * This function creates and allocate an array of futex_q (we zero it to
+ * initialize the fields) and then, for each futex_wait_block element from
+ * userspace, fill a futex_q element with proper values.
+ */
+inline struct futex_q *futex_read_wait_block(u32 __user *uaddr, u32 count)
+{
+ unsigned int i;
+ struct futex_q *qs;
+ struct futex_wait_block fwb;
+ struct futex_wait_block __user *entry =
+ (struct futex_wait_block __user *)uaddr;
+
+ if (!count || count > FUTEX_MULTIPLE_MAX_COUNT)
+ return ERR_PTR(-EINVAL);
+
+ qs = kcalloc(count, sizeof(*qs), GFP_KERNEL);
+ if (!qs)
+ return ERR_PTR(-ENOMEM);
+
+ for (i = 0; i < count; i++) {
+ if (copy_from_user(&fwb, &entry[i], sizeof(fwb))) {
+ kfree(qs);
+ return ERR_PTR(-EFAULT);
+ }
+
+ qs[i].uaddr = fwb.uaddr;
+ qs[i].uval = fwb.val;
+ qs[i].bitset = fwb.bitset;
+ }
+
+ return qs;
+}
SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val,
struct __kernel_timespec __user *, utime, u32 __user *, uaddr2,
@@ -3919,7 +4188,8 @@ SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val,
if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI ||
cmd == FUTEX_WAIT_BITSET ||
- cmd == FUTEX_WAIT_REQUEUE_PI)) {
+ cmd == FUTEX_WAIT_REQUEUE_PI ||
+ cmd == FUTEX_WAIT_MULTIPLE)) {
if (unlikely(should_fail_futex(!(op & FUTEX_PRIVATE_FLAG))))
return -EFAULT;
if (get_timespec64(&ts, utime))
@@ -3940,6 +4210,25 @@ SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val,
cmd == FUTEX_CMP_REQUEUE_PI || cmd == FUTEX_WAKE_OP)
val2 = (u32) (unsigned long) utime;
+ if (cmd == FUTEX_WAIT_MULTIPLE) {
+ int ret;
+ struct futex_q *qs;
+
+#ifdef CONFIG_X86_X32
+ if (unlikely(in_x32_syscall()))
+ return -ENOSYS;
+#endif
+ qs = futex_read_wait_block(uaddr, val);
+
+ if (IS_ERR(qs))
+ return PTR_ERR(qs);
+
+ ret = futex_wait_multiple(qs, op, val, tp);
+ kfree(qs);
+
+ return ret;
+ }
+
return do_futex(uaddr, op, val, tp, uaddr2, val2, val3);
}
@@ -4102,6 +4391,57 @@ COMPAT_SYSCALL_DEFINE3(get_robust_list, int, pid,
#endif /* CONFIG_COMPAT */
#ifdef CONFIG_COMPAT_32BIT_TIME
+/**
+ * struct compat_futex_wait_block - Block of futexes to be waited for
+ * @uaddr: User address of the futex (compatible pointer)
+ * @val: Futex value expected by userspace
+ * @bitset: Bitset for the optional bitmasked wakeup
+ */
+struct compat_futex_wait_block {
+ compat_uptr_t uaddr;
+ __u32 val;
+ __u32 bitset;
+};
+
+/**
+ * compat_futex_read_wait_block - Read an array of futex_wait_block from
+ * userspace
+ * @uaddr: Userspace address of the block
+ * @count: Number of blocks to be read
+ *
+ * This function does the same as futex_read_wait_block(), except that it
+ * converts the pointer to the futex from the compat version to the regular one.
+ */
+inline struct futex_q *compat_futex_read_wait_block(u32 __user *uaddr,
+ u32 count)
+{
+ unsigned int i;
+ struct futex_q *qs;
+ struct compat_futex_wait_block fwb;
+ struct compat_futex_wait_block __user *entry =
+ (struct compat_futex_wait_block __user *)uaddr;
+
+ if (!count || count > FUTEX_MULTIPLE_MAX_COUNT)
+ return ERR_PTR(-EINVAL);
+
+ qs = kcalloc(count, sizeof(*qs), GFP_KERNEL);
+ if (!qs)
+ return ERR_PTR(-ENOMEM);
+
+ for (i = 0; i < count; i++) {
+ if (copy_from_user(&fwb, &entry[i], sizeof(fwb))) {
+ kfree(qs);
+ return ERR_PTR(-EFAULT);
+ }
+
+ qs[i].uaddr = compat_ptr(fwb.uaddr);
+ qs[i].uval = fwb.val;
+ qs[i].bitset = fwb.bitset;
+ }
+
+ return qs;
+}
+
SYSCALL_DEFINE6(futex_time32, u32 __user *, uaddr, int, op, u32, val,
struct old_timespec32 __user *, utime, u32 __user *, uaddr2,
u32, val3)
@@ -4113,7 +4453,8 @@ SYSCALL_DEFINE6(futex_time32, u32 __user *, uaddr, int, op, u32, val,
if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI ||
cmd == FUTEX_WAIT_BITSET ||
- cmd == FUTEX_WAIT_REQUEUE_PI)) {
+ cmd == FUTEX_WAIT_REQUEUE_PI ||
+ cmd == FUTEX_WAIT_MULTIPLE)) {
if (get_old_timespec32(&ts, utime))
return -EFAULT;
if (!timespec64_valid(&ts))
@@ -4128,6 +4469,19 @@ SYSCALL_DEFINE6(futex_time32, u32 __user *, uaddr, int, op, u32, val,
cmd == FUTEX_CMP_REQUEUE_PI || cmd == FUTEX_WAKE_OP)
val2 = (int) (unsigned long) utime;
+ if (cmd == FUTEX_WAIT_MULTIPLE) {
+ int ret;
+ struct futex_q *qs = compat_futex_read_wait_block(uaddr, val);
+
+ if (IS_ERR(qs))
+ return PTR_ERR(qs);
+
+ ret = futex_wait_multiple(qs, op, val, tp);
+ kfree(qs);
+
+ return ret;
+ }
+
return do_futex(uaddr, op, val, tp, uaddr2, val2, val3);
}
#endif /* CONFIG_COMPAT_32BIT_TIME */
diff --git a/tools/testing/selftests/futex/functional/futex_wait_timeout.c b/tools/testing/selftests/futex/functional/futex_wait_timeout.c
index ee55e6d389a3f053194435342c4e471dc7cf8786..2a63e1c2cfb6407a5988233217cff2e52787bc66 100644
--- a/tools/testing/selftests/futex/functional/futex_wait_timeout.c
+++ b/tools/testing/selftests/futex/functional/futex_wait_timeout.c
@@ -11,6 +11,7 @@
*
* HISTORY
* 2009-Nov-6: Initial version by Darren Hart <dvhart@linux.intel.com>
+ * 2019-Dec-13: Add WAIT_MULTIPLE test by Krisman <krisman@collabora.com>
*
*****************************************************************************/
@@ -41,6 +42,8 @@ int main(int argc, char *argv[])
{
futex_t f1 = FUTEX_INITIALIZER;
struct timespec to;
+ time_t secs;
+ struct futex_wait_block fwb = {&f1, f1, 0};
int res, ret = RET_PASS;
int c;
@@ -65,7 +68,7 @@ int main(int argc, char *argv[])
}
ksft_print_header();
- ksft_set_plan(1);
+ ksft_set_plan(2);
ksft_print_msg("%s: Block on a futex and wait for timeout\n",
basename(argv[0]));
ksft_print_msg("\tArguments: timeout=%ldns\n", timeout_ns);
@@ -79,8 +82,39 @@ int main(int argc, char *argv[])
if (!res || errno != ETIMEDOUT) {
fail("futex_wait returned %d\n", ret < 0 ? errno : ret);
ret = RET_FAIL;
+ } else
+ ksft_test_result_pass("futex_wait timeout succeeds\n");
+
+ info("Calling futex_wait_multiple on f1: %u @ %p\n", f1, &f1);
+
+ /* Setup absolute time */
+ ret = clock_gettime(CLOCK_REALTIME, &to);
+ secs = (to.tv_nsec + timeout_ns) / 1000000000;
+ to.tv_nsec = ((int64_t)to.tv_nsec + timeout_ns) % 1000000000;
+ to.tv_sec += secs;
+ info("to.tv_sec = %ld\n", to.tv_sec);
+ info("to.tv_nsec = %ld\n", to.tv_nsec);
+
+ res = futex_wait_multiple(&fwb, 1, &to,
+ FUTEX_PRIVATE_FLAG | FUTEX_CLOCK_REALTIME);
+
+#ifdef __ILP32__
+ if (res == -1 && errno == ENOSYS) {
+ ksft_test_result_skip("futex_wait_multiple not supported at x32\n");
+ } else {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
}
+#else
+ if (!res || errno != ETIMEDOUT) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
+ } else
+ ksft_test_result_pass("futex_wait_multiple timeout succeeds\n");
+#endif /* __ILP32__ */
- print_result(TEST_NAME, ret);
+ ksft_print_cnts();
return ret;
}
diff --git a/tools/testing/selftests/futex/include/futextest.h b/tools/testing/selftests/futex/include/futextest.h
index ddbcfc9b7bac4aebb5bac2f249e26ecfd948aa84..bb103bef4557012ef9a389ca74c868e4476a8a31 100644
--- a/tools/testing/selftests/futex/include/futextest.h
+++ b/tools/testing/selftests/futex/include/futextest.h
@@ -38,6 +38,14 @@ typedef volatile u_int32_t futex_t;
#ifndef FUTEX_CMP_REQUEUE_PI
#define FUTEX_CMP_REQUEUE_PI 12
#endif
+#ifndef FUTEX_WAIT_MULTIPLE
+#define FUTEX_WAIT_MULTIPLE 13
+struct futex_wait_block {
+ futex_t *uaddr;
+ futex_t val;
+ __u32 bitset;
+};
+#endif
#ifndef FUTEX_WAIT_REQUEUE_PI_PRIVATE
#define FUTEX_WAIT_REQUEUE_PI_PRIVATE (FUTEX_WAIT_REQUEUE_PI | \
FUTEX_PRIVATE_FLAG)
@@ -80,6 +88,20 @@ futex_wait(futex_t *uaddr, futex_t val, struct timespec *timeout, int opflags)
return futex(uaddr, FUTEX_WAIT, val, timeout, NULL, 0, opflags);
}
+/**
+ * futex_wait_multiple() - block on several futexes with optional timeout
+ * @fwb: wait block user space address
+ * @count: number of entities at fwb
+ * @timeout: absolute timeout
+ */
+static inline int
+futex_wait_multiple(struct futex_wait_block *fwb, int count,
+ struct timespec *timeout, int opflags)
+{
+ return futex(fwb, FUTEX_WAIT_MULTIPLE, count, timeout, NULL, 0,
+ opflags);
+}
+
/**
* futex_wake() - wake one or more tasks blocked on uaddr
* @nr_wake: wake up to this many tasks
diff --git a/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c b/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
index 0ae390ff816449c88d0bb655a26eb014382c2b4f..bcbac042992d447e0bc9ef5fefe94e875de310f2 100644
--- a/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
+++ b/tools/testing/selftests/futex/functional/futex_wait_wouldblock.c
@@ -12,6 +12,7 @@
*
* HISTORY
* 2009-Nov-14: Initial version by Gowrishankar <gowrishankar.m@in.ibm.com>
+ * 2019-Dec-13: Add WAIT_MULTIPLE test by Krisman <krisman@collabora.com>
*
*****************************************************************************/
@@ -40,6 +41,7 @@ int main(int argc, char *argv[])
{
struct timespec to = {.tv_sec = 0, .tv_nsec = timeout_ns};
futex_t f1 = FUTEX_INITIALIZER;
+ struct futex_wait_block fwb = {&f1, f1+1, 0};
int res, ret = RET_PASS;
int c;
@@ -61,7 +63,7 @@ int main(int argc, char *argv[])
}
ksft_print_header();
- ksft_set_plan(1);
+ ksft_set_plan(2);
ksft_print_msg("%s: Test the unexpected futex value in FUTEX_WAIT\n",
basename(argv[0]));
@@ -71,8 +73,30 @@ int main(int argc, char *argv[])
fail("futex_wait returned: %d %s\n",
res ? errno : res, res ? strerror(errno) : "");
ret = RET_FAIL;
+ } else
+ ksft_test_result_pass("futex_wait wouldblock succeeds\n");
+
+ info("Calling futex_wait_multiple on f1: %u @ %p with val=%u\n",
+ f1, &f1, f1+1);
+ res = futex_wait_multiple(&fwb, 1, NULL, FUTEX_PRIVATE_FLAG);
+
+#ifdef __ILP32__
+ if (res != -1 || errno != ENOSYS) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
+ } else {
+ ksft_test_result_skip("futex_wait_multiple not supported at x32\n");
+ }
+#else
+ if (!res || errno != EWOULDBLOCK) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
}
+ ksft_test_result_pass("futex_wait_multiple wouldblock succeeds\n");
+#endif /* __ILP32__ */
- print_result(TEST_NAME, ret);
+ ksft_print_cnts();
return ret;
}
diff --git a/tools/testing/selftests/futex/functional/.gitignore b/tools/testing/selftests/futex/functional/.gitignore
index a09f570619023750f558c84004aff166b4337d72..4660128a545edb04a17cc6bd9760931c1386122f 100644
--- a/tools/testing/selftests/futex/functional/.gitignore
+++ b/tools/testing/selftests/futex/functional/.gitignore
@@ -5,3 +5,4 @@ futex_wait_private_mapped_file
futex_wait_timeout
futex_wait_uninitialized_heap
futex_wait_wouldblock
+futex_wait_multiple
diff --git a/tools/testing/selftests/futex/functional/Makefile b/tools/testing/selftests/futex/functional/Makefile
index 30996306cabcfe89a47977643e529b122893bb7e..75f9fface11fa3c90c1bdb9a49b3ea51291afd58 100644
--- a/tools/testing/selftests/futex/functional/Makefile
+++ b/tools/testing/selftests/futex/functional/Makefile
@@ -14,7 +14,8 @@ TEST_GEN_FILES := \
futex_requeue_pi_signal_restart \
futex_requeue_pi_mismatched_ops \
futex_wait_uninitialized_heap \
- futex_wait_private_mapped_file
+ futex_wait_private_mapped_file \
+ futex_wait_multiple
TEST_PROGS := run.sh
diff --git a/tools/testing/selftests/futex/functional/futex_wait_multiple.c b/tools/testing/selftests/futex/functional/futex_wait_multiple.c
new file mode 100644
index 0000000000000000000000000000000000000000..b48422e79f42edba1653bb0bd2a4c4fd98d2d48d
--- /dev/null
+++ b/tools/testing/selftests/futex/functional/futex_wait_multiple.c
@@ -0,0 +1,173 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/******************************************************************************
+ *
+ * Copyright © Collabora, Ltd., 2019
+ *
+ * DESCRIPTION
+ * Test basic semantics of FUTEX_WAIT_MULTIPLE
+ *
+ * AUTHOR
+ * Gabriel Krisman Bertazi <krisman@collabora.com>
+ *
+ * HISTORY
+ * 2019-Dec-13: Initial version by Krisman <krisman@collabora.com>
+ *
+ *****************************************************************************/
+
+#include <errno.h>
+#include <getopt.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include <pthread.h>
+#include "futextest.h"
+#include "logging.h"
+
+#define TEST_NAME "futex-wait-multiple"
+#define timeout_ns 100000
+#define MAX_COUNT 128
+#define WAKE_WAIT_US 3000000
+
+int ret = RET_PASS;
+char *progname;
+futex_t f[MAX_COUNT] = {0};
+struct futex_wait_block fwb[MAX_COUNT];
+
+void usage(char *prog)
+{
+ printf("Usage: %s\n", prog);
+ printf(" -c Use color\n");
+ printf(" -h Display this help message\n");
+ printf(" -v L Verbosity level: %d=QUIET %d=CRITICAL %d=INFO\n",
+ VQUIET, VCRITICAL, VINFO);
+}
+
+void test_count_overflow(void)
+{
+ futex_t f = FUTEX_INITIALIZER;
+ struct futex_wait_block fwb[MAX_COUNT+1];
+ int res, i;
+
+ ksft_print_msg("%s: Test a too big number of futexes\n", progname);
+
+ for (i = 0; i < MAX_COUNT+1; i++) {
+ fwb[i].uaddr = &f;
+ fwb[i].val = f;
+ fwb[i].bitset = 0;
+ }
+
+ res = futex_wait_multiple(fwb, MAX_COUNT+1, NULL, FUTEX_PRIVATE_FLAG);
+
+#ifdef __ILP32__
+ if (res != -1 || errno != ENOSYS) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
+ } else {
+ ksft_test_result_skip("futex_wait_multiple not supported at x32\n");
+ }
+#else
+ if (res != -1 || errno != EINVAL) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
+ } else {
+ ksft_test_result_pass("futex_wait_multiple count overflow succeed\n");
+ }
+
+#endif /* __ILP32__ */
+}
+
+void *waiterfn(void *arg)
+{
+ int res;
+
+ res = futex_wait_multiple(fwb, MAX_COUNT, NULL, FUTEX_PRIVATE_FLAG);
+
+#ifdef __ILP32__
+ if (res != -1 || errno != ENOSYS) {
+ ksft_test_result_fail("futex_wait_multiple returned %d\n",
+ res < 0 ? errno : res);
+ ret = RET_FAIL;
+ } else {
+ ksft_test_result_skip("futex_wait_multiple not supported at x32\n");
+ }
+#else
+ if (res < 0)
+ ksft_print_msg("waiter failed %d\n", res);
+
+ info("futex_wait_multiple: Got hint futex %d was freed\n", res);
+#endif /* __ILP32__ */
+
+ return NULL;
+}
+
+void test_fwb_wakeup(void)
+{
+ int res, i;
+ pthread_t waiter;
+
+ ksft_print_msg("%s: Test wake up in a list of futex\n", progname);
+
+ for (i = 0; i < MAX_COUNT; i++) {
+ fwb[i].uaddr = &f[i];
+ fwb[i].val = f[i];
+ fwb[i].bitset = 0xffffffff;
+ }
+
+ res = pthread_create(&waiter, NULL, waiterfn, NULL);
+ if (res) {
+ ksft_test_result_fail("Creating waiting thread failed");
+ ksft_exit_fail();
+ }
+
+ usleep(WAKE_WAIT_US);
+ res = futex_wake(&(f[MAX_COUNT-1]), 1, FUTEX_PRIVATE_FLAG);
+ if (res != 1) {
+ ksft_test_result_fail("Failed to wake thread res=%d\n", res);
+ ksft_exit_fail();
+ }
+
+ pthread_join(waiter, NULL);
+ ksft_test_result_pass("%s succeed\n", __func__);
+}
+
+int main(int argc, char *argv[])
+{
+ int c;
+
+ while ((c = getopt(argc, argv, "cht:v:")) != -1) {
+ switch (c) {
+ case 'c':
+ log_color(1);
+ break;
+ case 'h':
+ usage(basename(argv[0]));
+ exit(0);
+ case 'v':
+ log_verbosity(atoi(optarg));
+ break;
+ default:
+ usage(basename(argv[0]));
+ exit(1);
+ }
+ }
+
+ progname = basename(argv[0]);
+
+ ksft_print_header();
+ ksft_set_plan(2);
+
+ test_count_overflow();
+
+#ifdef __ILP32__
+ // if it's a 32x binary, there's no futex to wakeup
+ ksft_test_result_skip("futex_wait_multiple not supported at x32\n");
+#else
+ test_fwb_wakeup();
+#endif /* __ILP32__ */
+
+ ksft_print_cnts();
+ return ret;
+}
diff --git a/tools/testing/selftests/futex/functional/run.sh b/tools/testing/selftests/futex/functional/run.sh
index 1acb6ace1680e8f3d6b3ee2dc528c19ddfdb018e..a8be94f28ff78b4879d2d19bca5d9b0fcb26c1f8 100755
--- a/tools/testing/selftests/futex/functional/run.sh
+++ b/tools/testing/selftests/futex/functional/run.sh
@@ -73,3 +73,6 @@ echo
echo
./futex_wait_uninitialized_heap $COLOR
./futex_wait_private_mapped_file $COLOR
+
+echo
+./futex_wait_multiple $COLOR
diff --git a/include/uapi/linux/futex.h b/include/uapi/linux/futex.h
index 580001e89c6caed57dd8b3cb491d65dce846caff..a3e760886b8e7e74285fdcf2caaaa6f66ad16675 100644
--- a/include/uapi/linux/futex.h
+++ b/include/uapi/linux/futex.h
@@ -21,7 +21,7 @@
#define FUTEX_WAKE_BITSET 10
#define FUTEX_WAIT_REQUEUE_PI 11
#define FUTEX_CMP_REQUEUE_PI 12
-#define FUTEX_WAIT_MULTIPLE 13
+#define FUTEX_WAIT_MULTIPLE 31
#define FUTEX_PRIVATE_FLAG 128
#define FUTEX_CLOCK_REALTIME 256
diff --git a/kernel/futex.c b/kernel/futex.c
index 58cf9eb2b851b4858e29b5ef4114a29a92e676ba..e0bb628a5e1988dcc9ae5442a4259edc229d578d 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -4198,7 +4198,7 @@ SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val,
return -EINVAL;
t = timespec64_to_ktime(ts);
- if (cmd == FUTEX_WAIT)
+ if (cmd == FUTEX_WAIT || cmd == FUTEX_WAIT_MULTIPLE)
t = ktime_add_safe(ktime_get(), t);
tp = &t;
}
@@ -4399,6 +4399,7 @@ COMPAT_SYSCALL_DEFINE3(get_robust_list, int, pid,
*/
struct compat_futex_wait_block {
compat_uptr_t uaddr;
+ __u32 pad;
__u32 val;
__u32 bitset;
};
@@ -4461,7 +4462,7 @@ SYSCALL_DEFINE6(futex_time32, u32 __user *, uaddr, int, op, u32, val,
return -EINVAL;
t = timespec64_to_ktime(ts);
- if (cmd == FUTEX_WAIT)
+ if (cmd == FUTEX_WAIT || cmd == FUTEX_WAIT_MULTIPLE)
t = ktime_add_safe(ktime_get(), t);
tp = &t;
}

43
linux58-rc-tkg/linux58-tkg-patches/0011-ZFS-fix.patch Normal file
View File

@ -0,0 +1,43 @@
From 1e010beda2896bdf3082fb37a3e49f8ce20e04d8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?J=C3=B6rg=20Thalheim?= <joerg@thalheim.io>
Date: Thu, 2 May 2019 05:28:08 +0100
Subject: [PATCH] x86/fpu: Export kernel_fpu_{begin,end}() with
EXPORT_SYMBOL_GPL
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
We need these symbols in zfs as the fpu implementation breaks userspace:
https://github.com/zfsonlinux/zfs/issues/9346
Signed-off-by: Jörg Thalheim <joerg@thalheim.io>
---
arch/x86/kernel/fpu/core.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index 12c70840980e..352538b3bb5d 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -102,7 +102,7 @@ void kernel_fpu_begin(void)
}
__cpu_invalidate_fpregs_state();
}
-EXPORT_SYMBOL_GPL(kernel_fpu_begin);
+EXPORT_SYMBOL(kernel_fpu_begin);
void kernel_fpu_end(void)
{
@@ -111,7 +111,7 @@ void kernel_fpu_end(void)
this_cpu_write(in_kernel_fpu, false);
preempt_enable();
}
-EXPORT_SYMBOL_GPL(kernel_fpu_end);
+EXPORT_SYMBOL(kernel_fpu_end);
/*
* Save the FPU state (mark it for reload if necessary):
--
2.23.0