Archlinux Kernel based on different schedulers and some other performance improvements.
Shell archlinuxbaby-schedcachycachy-schedulercachyoscacule-schedkernellinux-kernelperformanceperformance-tuning
Use environment variables to override parameters in PKGBUILD
Vasiliy Stelmachenok2cfbfb122bd0
diff --git a/linux-bore/PKGBUILD b/linux-bore/PKGBUILDindex 7970e395..f61bddd8 100644--- a/linux-bore/PKGBUILD+++ b/linux-bore/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='bore'+_cpusched=${_cpusched='bore'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=${_cpusched}-ltodiff --git a/linux-cachyos-bmq/PKGBUILD b/linux-cachyos-bmq/PKGBUILDindex c32f27da..4bcbd8e3 100644--- a/linux-cachyos-bmq/PKGBUILD+++ b/linux-cachyos-bmq/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='bmq'+_cpusched=${_cpusched='bmq'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-bore/PKGBUILD b/linux-cachyos-bore/PKGBUILDindex 4a124013..54850ee9 100644--- a/linux-cachyos-bore/PKGBUILD+++ b/linux-cachyos-bore/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='bore'+_cpusched=${_cpusched='bore'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-cacule/PKGBUILD b/linux-cachyos-cacule/PKGBUILDindex dd766f18..3575f7d0 100644--- a/linux-cachyos-cacule/PKGBUILD+++ b/linux-cachyos-cacule/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='cacule'+_cpusched=${_cpusched='cacule'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-cfs/PKGBUILD b/linux-cachyos-cfs/PKGBUILDindex 04eb7029..3fb1d857 100644--- a/linux-cachyos-cfs/PKGBUILD+++ b/linux-cachyos-cfs/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='cfs'+_cpusched=${_cpusched='cfs'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-hardened/PKGBUILD b/linux-cachyos-hardened/PKGBUILDindex ba6e1a56..840150a9 100644--- a/linux-cachyos-hardened/PKGBUILD+++ b/linux-cachyos-hardened/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='hardened'+_cpusched=${_cpusched='hardened'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-pds/PKGBUILD b/linux-cachyos-pds/PKGBUILDindex 2a38f75e..0dd42172 100644--- a/linux-cachyos-pds/PKGBUILD+++ b/linux-cachyos-pds/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='pds'+_cpusched=${_cpusched='pds'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos-rc/PKGBUILD b/linux-cachyos-rc/PKGBUILDindex 25d4dc5f..76690f92 100644--- a/linux-cachyos-rc/PKGBUILD+++ b/linux-cachyos-rc/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='bore'+_cpusched=${_cpusched='bore'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,16 +147,16 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable RT kernel-_rtkernel=+_rtkernel=${_rtkernel=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -166,7 +166,7 @@ _rtkernel=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attributediff --git a/linux-cachyos-tt/PKGBUILD b/linux-cachyos-tt/PKGBUILDindex 97c62360..8690c77c 100644--- a/linux-cachyos-tt/PKGBUILD+++ b/linux-cachyos-tt/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='tt'+_cpusched=${_cpusched='tt'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-${_cpusched}-ltodiff --git a/linux-cachyos/PKGBUILD b/linux-cachyos/PKGBUILDindex 7a2ec2a1..bbbc2507 100644--- a/linux-cachyos/PKGBUILD+++ b/linux-cachyos/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='bore'+_cpusched=${_cpusched='bore'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=cachyos-ltodiff --git a/linux-cacule-rdb/PKGBUILD b/linux-cacule-rdb/PKGBUILDindex 4b5c170e..2ab0606e 100644--- a/linux-cacule-rdb/PKGBUILD+++ b/linux-cacule-rdb/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='cacule-rdb'+_cpusched=${_cpusched='cacule-rdb'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=${_cpusched}-ltodiff --git a/linux-cacule/PKGBUILD b/linux-cacule/PKGBUILDindex bb2be0a2..d5d83f63 100644--- a/linux-cacule/PKGBUILD+++ b/linux-cacule/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='cacule'+_cpusched=${_cpusched='cacule'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=${_cpusched}-ltodiff --git a/linux-tt/PKGBUILD b/linux-tt/PKGBUILDindex 34d17a2f..717adc2d 100644--- a/linux-tt/PKGBUILD+++ b/linux-tt/PKGBUILD@@ -7,7 +7,7 @@# ATTENTION - one of two predefined values should be selected!# 'yes' - enable CachyOS config# 'no' - disable CachyOS config-_cachy_config='yes'+_cachy_config=${_cachy_config='yes'}### Selecting the CPU scheduler# ATTENTION - one of seven predefined values should be selected!@@ -19,27 +19,27 @@ _cachy_config='yes'# 'cfs' - select 'Completely Fair Scheduler'# 'tt' - select 'Task Type Scheduler by Hamad Marri'# 'hardened' - select 'BORE Scheduler hardened' ## kernel with hardened config and hardening patches with the bore scheduler-_cpusched='tt'+_cpusched=${_cpusched='tt'}### BUILD OPTIONS# Set these variables to ANYTHING that is not null to enable them### Tweak kernel options prior to a build via nconfig-_makenconfig=+_makenconfig=${_makenconfig=}### Tweak kernel options prior to a build via menuconfig-_makemenuconfig=+_makemenuconfig=${_makemenuconfig=}### Tweak kernel options prior to a build via xconfig-_makexconfig=+_makexconfig=${_makexconfig=}### Tweak kernel options prior to a build via gconfig-_makegconfig=+_makegconfig=${_makegconfig=}# NUMA is optimized for multi-socket motherboards.# A single multi-core CPU actually runs slower with NUMA enabled.# See, https://bugs.archlinux.org/task/31187-_NUMAdisable=y+_NUMAdisable=${_NUMAdisable=y}# Compile ONLY used modules to VASTLYreduce the number of modules built# and the build time.@@ -49,62 +49,62 @@ _NUMAdisable=y# This PKGBUILD read the database kept if it exists## More at this wiki page ---> https://wiki.archlinux.org/index.php/Modprobed-db-_localmodcfg=+_localmodcfg=${_localmodcfg=}# Use the current kernel's .config file# Enabling this option will use the .config of the RUNNING kernel rather than# the ARCH defaults. Useful when the package gets updated and you already went# through the trouble of customizing your config options. NOT recommended when# a new kernel is released, but again, convenient for package bumps.-_use_current=+_use_current=${_use_current=}### Enable KBUILD_CFLAGS -O3-_cc_harder=y+_cc_harder=${_cc_harder=y}### Set this to your number of threads you have in your machine otherwise it will default to 128-_nr_cpus=+_nr_cpus=${_nr_cpus=}### Set performance governor as default-_per_gov=y+_per_gov=${_per_gov=y}### Enable TCP_CONG_BBR2-_tcp_bbr2=y+_tcp_bbr2=${_tcp_bbr2=y}### Running with a 1000HZ, 750Hz, 600 Hz, 500Hz, 300Hz, 250Hz and 100Hz tick rate-_HZ_ticks=750+_HZ_ticks=${_HZ_ticks=750}## Choose between perodic, idle or full### Full tickless can give higher performances in various cases but, depending on hardware, lower consistency.-_tickrate=full+_tickrate=${_tickrate=full}## Choose between full(low-latency), voluntary or server-_preempt=full+_preempt=${_preempt=full}### Disable MQ-Deadline I/O scheduler-_mq_deadline_disable=y+_mq_deadline_disable=${_mq_deadline_disable=y}### Disable Kyber I/O scheduler-_kyber_disable=y+_kyber_disable=${_kyber_disable=y}### Enable multigenerational LRU# ATTENTION - one of three predefined values should be selected!# 'standard' - enable multigenerational LRU# 'stats' - enable multigenerational LRU with stats# 'none' - disable multigenerational LRU-_lru_config='standard'+_lru_config=${_lru_config='standard'}### Enable per-VMA locking# ATTENTION - one of three predefined values should be selected!# 'standard' - enable per-VMA locking# 'stats' - enable per-VMA locking with stats# 'none' - disable per-VMA locking-_vma_config='standard'+_vma_config=${_vma_config='standard'}## Enable DAMON-_damon=+_damon=${_damon=}## Enable Linux Random Number Generator-_lrng_enable=y+_lrng_enable=${_lrng_enable=y}# CPU compiler optimizations - Defaults to prompt at kernel config if left empty# AMD CPUs : "k8" "k8sse3" "k10" "barcelona" "bobcat" "jaguar" "bulldozer" "piledriver" "steamroller" "excavator" "zen" "zen2" "zen3"@@ -115,15 +115,15 @@ _lrng_enable=y# - "generic" (kernel's default - to share the package between machines with different CPU µarch as long as they are x86-64)## Or use the _use_auto_optimization with _use_auto_optimization=y-_processor_opt=+_processor_opt=${_processor_opt=}-_use_auto_optimization=y+_use_auto_optimization=${_use_auto_optimization=y}# disable debug to lower the size of the kernel-_disable_debug=y+_disable_debug=${_disable_debug=y}## Enable zram/zswap ZSTD compression-_zstd_compression=y+_zstd_compression=${_zstd_compression=y}### Selecting the ZSTD kernel and modules compression level# ATTENTION - one of two predefined values should be selected!@@ -131,12 +131,12 @@ _zstd_compression=y# 'normal' - standard compression ratio# WARNING: the ultra settings can sometimes# be counterproductive in both size and speed.-_zstd_level_value='normal'+_zstd_level_value=${_zstd_level_value='normal'}# Clang LTO mode, only available with the "llvm" compiler - options are "no", "full" or "thin".# "full: uses 1 thread for Linking, slow and uses more memory, theoretically with the highest performance gains."# "thin: uses multiple threads, faster and uses less memory, may have a lower runtime performance than Full."-_use_llvm_lto=+_use_llvm_lto=${_use_llvm_lto=}# KCFI is a proposed forward-edge control-flow integrity scheme for# Clang, which is more suitable for kernel use than the existing CFI@@ -147,13 +147,13 @@ _use_llvm_lto=# you can find a patched llvm-git in the cachyos-repo's.# The packagename is called "llvm-kcfi"# ATTENTION!: This is very experimental and could fail and the compilation or have other bugs in the kernel-_use_kcfi=+_use_kcfi=${_use_kcfi=}# Build the zfs module builtin in to the kernel-_build_zfs=+_build_zfs=${_build_zfs=}# Enable bcachefs-_bcachefs=+_bcachefs=${_bcachefs=}# Enable NEST# NEST is a experimental cfs scheduler you can find more about here:@@ -163,7 +163,7 @@ _bcachefs=# taskset -c $THREADS application# example: taskset -c 0-23 application# ATTENTION!:Just works together with the BORE Scheduler and CFS Scheduler-_nest=+_nest=${_nest=}# Enable LATENCY NICE# Latency nice is a approach to sets latency-nice as a per-task attribute@@ -174,7 +174,7 @@ _nest=# You need to configure ananicy-cpp for this or use existing settings# If you want to test it, use the following branch# https://gitlab.com/ananicy-cpp/ananicy-cpp/-/tree/feature/latency-nice-_latency_nice=+_latency_nice=${_latency_nice=}if [ -n "$_use_llvm_lto" ]; thenpkgsuffix=${_cpusched}-lto