testsAndMisc/linux_configuration/scripts/single_use/fixes/optimize_arch_desktop.sh
Krzysztof kuhy Rudnicki 42a66a1419 refactor(linux_configuration/scripts): split all scripts into single_use/ and periodic_background/
- Move all linux_configuration scripts into two semantic categories:
  - single_use/: scripts run once manually (fresh install, fixes, setup)
  - periodic_background/: scripts run by systemd timers or daemons
- Preserve existing subdirectory structure within each category
- Fix lib/common.sh source paths for new directory depths
- Fix CONFIG_DIR depth in setup_periodic_system.sh and check_and_enable_services.sh
- Update all references in tests, fresh-install/main.sh, nix modules, and docs
- Fix check_polling_antipatterns.sh false positives (||, regex |, case patterns, jq strings)
- Fix pre-existing mypy exclusion path and type annotations for moved tools/ directory
- Rewrite check_polling_antipatterns.sh using awk (no bash regex loops); add require_serial: true
2026-05-15 00:32:35 +02:00

624 lines
19 KiB
Bash
Executable File

#!/usr/bin/env bash
# Optimize Arch Linux desktop for maximum performance on high-end hardware
#
# Tuning areas:
# 1. CPU scheduler — performance governor on all cores
# 2. I/O scheduler — optimal scheduler per drive type (none for NVMe, mq-deadline for SATA SSD)
# 3. Memory / swap — lower swappiness, tune dirty page writeback for responsiveness
# 4. Kernel network — TCP BBR, fastopen, larger buffers
# 5. Filesystem — fstrim timer, noatime advisory
# 6. NVIDIA GPU — max performance level via persistence mode
# 7. Kernel mitigations — option to disable CPU vulnerability mitigations for extra speed
# 8. Boot speed — disable unnecessary wait-online services
# 9. Journal housekeeping — cap at 300M
# 10. Process scheduler — install ananicy-cpp for automatic nice/ionice/scheduling
#
# Usage:
# ./optimize_arch_desktop.sh # Apply safe optimizations
# ./optimize_arch_desktop.sh --dry-run # Show what would be done
# ./optimize_arch_desktop.sh --interactive # Prompt before each tweak
# ./optimize_arch_desktop.sh --aggressive # Include CPU mitigation disable (risk: security)
# ./optimize_arch_desktop.sh -h # Show help
#
# All tweaks are idempotent and safe to re-run.
# Some kernel parameter changes require a reboot to take full effect.
set -euo pipefail
SCRIPT_DIR="$(dirname "$(readlink -f "$0")")"
# shellcheck source=../../lib/common.sh
source "$SCRIPT_DIR/../../lib/common.sh"
parse_interactive_args "$@"
shift "$COMMON_ARGS_SHIFT"
DRY_RUN=false
AGGRESSIVE=false
for arg in "$@"; do
case "$arg" in
--dry-run)
DRY_RUN=true
;;
--aggressive)
AGGRESSIVE=true
;;
-h | --help)
cat <<'EOF'
optimize_arch_desktop.sh - Squeeze maximum performance from an Arch Linux desktop
Usage: optimize_arch_desktop.sh [OPTIONS]
Options:
--dry-run Show what would be done without making changes
--aggressive Also disable CPU vulnerability mitigations (trades security for speed)
-i, --interactive Prompt before each optimization
-h, --help Show this help message
Optimizations applied:
1. Set CPU governor to performance on all cores
2. Set optimal I/O scheduler per drive (none/mq-deadline)
3. Tune vm.swappiness, dirty ratios, vfs_cache_pressure via sysctl
4. Enable TCP BBR congestion control + fastopen + buffer tuning
5. Enable fstrim.timer for SSD TRIM maintenance
6. Set NVIDIA GPU to max performance level (persistence mode)
7. [--aggressive] Disable CPU vulnerability mitigations
8. Disable NetworkManager-wait-online.service for faster boot
9. Vacuum & cap systemd journal at 300M
10. Install/enable ananicy-cpp for automatic process prioritization
All optimizations are idempotent. Re-run safely at any time.
EOF
exit 0
;;
esac
done
require_root "$@"
print_setup_header "Arch Linux Desktop Performance Optimizer"
TWEAKS_APPLIED=0
TWEAKS_SKIPPED=0
# ---------------------------------------------------------------------------
# Helper: apply or preview a tweak
# ---------------------------------------------------------------------------
apply_tweak() {
local description="$1"
shift
echo ""
log_info "$description"
if [[ $DRY_RUN == "true" ]]; then
echo " [dry-run] Would run: $*"
return 0
fi
if [[ $INTERACTIVE_MODE == "true" ]]; then
if ! ask_yes_no " Apply this optimization?"; then
log_warn "Skipped."
((TWEAKS_SKIPPED++)) || true
return 0
fi
fi
if "$@"; then
log_ok "Done."
((TWEAKS_APPLIED++)) || true
else
log_error "Failed (non-fatal, continuing)."
fi
}
# ===================================================================
# 1. CPU Governor → performance
# ===================================================================
tweak_cpu_governor() {
local gov_files
gov_files=$(find /sys/devices/system/cpu -maxdepth 3 -name scaling_governor 2>/dev/null || true)
if [[ -z $gov_files ]]; then
log_warn "No CPU governor sysfs files found — skipping."
return 0
fi
# Check current state
local all_performance=true
local f
for f in $gov_files; do
if [[ $(cat "$f") != "performance" ]]; then
all_performance=false
break
fi
done
if [[ $all_performance == "true" ]]; then
log_ok "All CPU cores already on 'performance' governor — skipping."
return 0
fi
for f in $gov_files; do
echo "performance" >"$f"
done
# Make it persistent via a sysctl-style drop-in using udev rule
local udev_rule="/etc/udev/rules.d/60-cpu-governor-performance.rules"
if [[ ! -f $udev_rule ]]; then
cat >"$udev_rule" <<'UDEVEOF'
# Set CPU governor to performance on all cores at boot
SUBSYSTEM=="module", DEVPATH=="*/cpu/*", ATTR{scaling_governor}=="*", ATTR{scaling_governor}="performance"
UDEVEOF
fi
# Also install cpupower hook as a more reliable persistence method
local cpupower_conf="/etc/default/cpupower"
if has_cmd cpupower; then
if [[ ! -f $cpupower_conf ]] || ! grep -q "^governor='performance'" "$cpupower_conf" 2>/dev/null; then
mkdir -p "$(dirname "$cpupower_conf")"
cat >"$cpupower_conf" <<'CPUEOF'
# /etc/default/cpupower — managed by optimize_arch_desktop.sh
governor='performance'
CPUEOF
systemctl enable cpupower.service 2>/dev/null || true
fi
fi
return 0
}
# ===================================================================
# 2. I/O Scheduler per drive type
# ===================================================================
tweak_io_scheduler() {
local changed=false
local block_dev
for block_dev in /sys/block/sd* /sys/block/nvme* /sys/block/vd*; do
[[ -d $block_dev ]] || continue
local sched_file="$block_dev/queue/scheduler"
[[ -f $sched_file ]] || continue
local dev_name
dev_name=$(basename "$block_dev")
local rotational
rotational=$(cat "$block_dev/queue/rotational" 2>/dev/null || echo 1)
local current
current=$(sed 's/.*\[\(.*\)\].*/\1/' "$sched_file" 2>/dev/null || true)
local target
if [[ $dev_name == nvme* ]]; then
target="none"
elif [[ $rotational -eq 0 ]]; then
target="mq-deadline"
else
target="bfq"
fi
if [[ $current == "$target" ]]; then
log_ok "$dev_name: already using '$target' scheduler."
continue
fi
echo "$target" >"$sched_file" 2>/dev/null || true
log_info "$dev_name: scheduler changed from '$current' to '$target'."
changed=true
done
# Persist via udev rule
local udev_rule="/etc/udev/rules.d/60-io-scheduler.rules"
if [[ ! -f $udev_rule ]]; then
cat >"$udev_rule" <<'IOEOF'
# NVMe: no scheduler (multi-queue hardware handles it)
ACTION=="add|change", KERNEL=="nvme[0-9]*", ATTR{queue/scheduler}="none"
# SATA SSD: mq-deadline (low latency)
ACTION=="add|change", KERNEL=="sd[a-z]", ATTR{queue/rotational}=="0", ATTR{queue/scheduler}="mq-deadline"
# HDD: BFQ (fair bandwidth allocation)
ACTION=="add|change", KERNEL=="sd[a-z]", ATTR{queue/rotational}=="1", ATTR{queue/scheduler}="bfq"
IOEOF
fi
if [[ $changed == "false" ]]; then
log_ok "All I/O schedulers already optimal."
fi
return 0
}
# ===================================================================
# 3. Memory & swap tuning via sysctl
# ===================================================================
tweak_vm_sysctl() {
local dropin="/etc/sysctl.d/90-desktop-performance.conf"
# Desktop workloads: low swappiness, aggressive VFS caching, tuned dirty ratios
local -A params=(
["vm.swappiness"]="10"
["vm.vfs_cache_pressure"]="50"
["vm.dirty_ratio"]="15"
["vm.dirty_background_ratio"]="5"
["vm.dirty_writeback_centisecs"]="1500"
["vm.page-cluster"]="0"
)
local needs_update=false
local key
for key in "${!params[@]}"; do
local current
current=$(sysctl -n "$key" 2>/dev/null || true)
if [[ $current != "${params[$key]}" ]]; then
needs_update=true
break
fi
done
if [[ $needs_update == "false" && -f $dropin ]]; then
log_ok "VM sysctl parameters already tuned — skipping."
return 0
fi
cat >"$dropin" <<'VMEOF'
# Desktop performance tuning — managed by optimize_arch_desktop.sh
#
# vm.swappiness=10 — prefer keeping data in RAM over swapping
# vm.vfs_cache_pressure=50 — favor keeping inode/dentry caches (speeds up file operations)
# vm.dirty_ratio=15 — allow up to 15% RAM dirty before synchronous writeback
# vm.dirty_background_ratio=5 — start async writeback at 5% dirty
# vm.dirty_writeback_centisecs=1500 — flush dirty pages every 15s (less I/O churn)
# vm.page-cluster=0 — read one page at a time from swap (reduces latency on SSD)
vm.swappiness = 10
vm.vfs_cache_pressure = 50
vm.dirty_ratio = 15
vm.dirty_background_ratio = 5
vm.dirty_writeback_centisecs = 1500
vm.page-cluster = 0
VMEOF
sysctl --system >/dev/null 2>&1
return 0
}
# ===================================================================
# 4. Network: TCP BBR + fastopen + buffer tuning
# ===================================================================
tweak_network_sysctl() {
local dropin="/etc/sysctl.d/91-desktop-network.conf"
# Check if BBR module is available
if ! modprobe tcp_bbr 2>/dev/null; then
log_warn "tcp_bbr kernel module unavailable — skipping network tuning."
return 0
fi
local -A params=(
["net.core.default_qdisc"]="fq"
["net.ipv4.tcp_congestion_control"]="bbr"
["net.ipv4.tcp_fastopen"]="3"
["net.core.rmem_max"]="16777216"
["net.core.wmem_max"]="16777216"
["net.ipv4.tcp_rmem"]="4096 1048576 16777216"
["net.ipv4.tcp_wmem"]="4096 1048576 16777216"
["net.ipv4.tcp_mtu_probing"]="1"
)
local needs_update=false
local key
for key in "${!params[@]}"; do
local current
current=$(sysctl -n "$key" 2>/dev/null || true)
# Normalize whitespace for comparison (kernel uses tabs)
current=$(echo "$current" | xargs)
local expected
expected=$(echo "${params[$key]}" | xargs)
if [[ $current != "$expected" ]]; then
needs_update=true
break
fi
done
if [[ $needs_update == "false" && -f $dropin ]]; then
log_ok "Network sysctl parameters already tuned — skipping."
return 0
fi
cat >"$dropin" <<'NETEOF'
# Network performance tuning — managed by optimize_arch_desktop.sh
#
# BBR congestion control — better throughput and lower latency than cubic
# TCP fastopen — saves one RTT on repeated connections (both client and server)
# Larger buffers — helps on high-bandwidth or high-latency links
net.core.default_qdisc = fq
net.ipv4.tcp_congestion_control = bbr
net.ipv4.tcp_fastopen = 3
net.core.rmem_max = 16777216
net.core.wmem_max = 16777216
net.ipv4.tcp_rmem = 4096 1048576 16777216
net.ipv4.tcp_wmem = 4096 1048576 16777216
net.ipv4.tcp_mtu_probing = 1
NETEOF
sysctl --system >/dev/null 2>&1
return 0
}
# ===================================================================
# 5. fstrim timer
# ===================================================================
tweak_fstrim() {
if systemctl is-enabled fstrim.timer >/dev/null 2>&1; then
log_ok "fstrim.timer already enabled — skipping."
return 0
fi
systemctl enable --now fstrim.timer
return 0
}
# ===================================================================
# 6. NVIDIA GPU — max performance
# ===================================================================
tweak_nvidia_gpu() {
if ! has_cmd nvidia-smi; then
log_info "nvidia-smi not found — skipping GPU tuning."
return 0
fi
# Enable persistence mode (keeps driver loaded, faster app launches)
local persist_status
persist_status=$(nvidia-smi --query-gpu=persistence_mode --format=csv,noheader 2>/dev/null | head -n 1 | xargs || true)
if [[ $persist_status != "Enabled" ]]; then
nvidia-smi -pm 1 >/dev/null 2>&1 || true
log_info "NVIDIA persistence mode enabled."
else
log_ok "NVIDIA persistence mode already enabled."
fi
# Set power management to prefer maximum performance
# PowerMizerMode: 1 = prefer max perf
nvidia-smi -gps 0 >/dev/null 2>&1 || true
# Persist via systemd service
local service_file="/etc/systemd/system/nvidia-performance.service"
if [[ ! -f $service_file ]]; then
cat >"$service_file" <<'NVSVC'
[Unit]
Description=Set NVIDIA GPU to max performance mode
After=nvidia-persistenced.service
[Service]
Type=oneshot
ExecStart=/usr/bin/nvidia-smi -pm 1
RemainAfterExit=yes
[Install]
WantedBy=multi-user.target
NVSVC
systemctl daemon-reload
systemctl enable nvidia-performance.service 2>/dev/null || true
fi
# Ensure nvidia-persistenced is enabled
if has_cmd nvidia-persistenced; then
systemctl enable nvidia-persistenced.service 2>/dev/null || true
if ! systemctl is-active nvidia-persistenced.service >/dev/null 2>&1; then
systemctl start nvidia-persistenced.service 2>/dev/null || true
fi
fi
return 0
}
# ===================================================================
# 7. [AGGRESSIVE] Disable CPU vulnerability mitigations
# ===================================================================
tweak_mitigations() {
if [[ $AGGRESSIVE != "true" ]]; then
log_info "Skipping CPU mitigation disable (use --aggressive to enable)."
return 0
fi
# Detect boot loader
local boot_method=""
if [[ -d /boot/loader/entries ]]; then
boot_method="systemd-boot"
elif [[ -f /etc/default/grub ]]; then
boot_method="grub"
else
log_warn "Could not detect boot loader — skipping mitigation tweak."
log_info "Manually add 'mitigations=off' to your kernel command line for extra speed."
return 0
fi
if [[ $boot_method == "systemd-boot" ]]; then
local entry
entry=$(find /boot/loader/entries -name '*.conf' -print -quit 2>/dev/null || true)
if [[ -n $entry ]]; then
if grep -q 'mitigations=off' "$entry" 2>/dev/null; then
log_ok "mitigations=off already set in systemd-boot — skipping."
return 0
fi
# Append to the options line
sed -i '/^options / s/$/ mitigations=off/' "$entry"
log_warn "Added mitigations=off to $entry. REBOOT REQUIRED."
log_warn "This trades security for ~5-15% performance. Only for isolated desktops."
fi
elif [[ $boot_method == "grub" ]]; then
if grep -q 'mitigations=off' /etc/default/grub 2>/dev/null; then
log_ok "mitigations=off already set in GRUB — skipping."
return 0
fi
sed -i 's/^GRUB_CMDLINE_LINUX_DEFAULT="\(.*\)"/GRUB_CMDLINE_LINUX_DEFAULT="\1 mitigations=off"/' /etc/default/grub
grub-mkconfig -o /boot/grub/grub.cfg
log_warn "Added mitigations=off to GRUB config. REBOOT REQUIRED."
fi
return 0
}
# ===================================================================
# 8. Disable NetworkManager-wait-online
# ===================================================================
tweak_nm_wait_online() {
if ! systemctl is-enabled NetworkManager-wait-online.service >/dev/null 2>&1; then
log_ok "NetworkManager-wait-online already disabled — skipping."
return 0
fi
systemctl disable NetworkManager-wait-online.service
return 0
}
# ===================================================================
# 9. Journal vacuum + permanent cap
# ===================================================================
tweak_journal() {
local usage_line
usage_line=$(journalctl --disk-usage 2>/dev/null || true)
local needs_vacuum=false
if [[ $usage_line =~ ([0-9]+\.?[0-9]*)\ G ]]; then
needs_vacuum=true
fi
if [[ $needs_vacuum == "true" ]]; then
journalctl --vacuum-size=300M
else
log_ok "Journal already under 1GiB."
fi
local dropin_dir="/etc/systemd/journald.conf.d"
local dropin_file="$dropin_dir/size-limit.conf"
if [[ -f $dropin_file ]] && grep -q 'SystemMaxUse=300M' "$dropin_file"; then
log_ok "Journal size cap already configured."
else
mkdir -p "$dropin_dir"
cat >"$dropin_file" <<'JOURNALEOF'
[Journal]
SystemMaxUse=300M
JOURNALEOF
systemctl restart systemd-journald
fi
return 0
}
# ===================================================================
# 10. ananicy-cpp — automatic process nice/ionice/scheduler tuning
# ===================================================================
tweak_ananicy() {
# ananicy-cpp is the C++ rewrite, available in the AUR via ananicy-cpp
if systemctl is-enabled ananicy-cpp.service >/dev/null 2>&1; then
log_ok "ananicy-cpp is already enabled — skipping."
return 0
fi
if pacman -Qi ananicy-cpp >/dev/null 2>&1; then
systemctl enable --now ananicy-cpp.service
log_info "Enabled ananicy-cpp.service."
return 0
fi
# Check for the original ananicy
if pacman -Qi ananicy >/dev/null 2>&1; then
if ! systemctl is-enabled ananicy.service >/dev/null 2>&1; then
systemctl enable --now ananicy.service
log_info "Enabled ananicy.service."
else
log_ok "ananicy is already enabled."
fi
return 0
fi
log_info "ananicy-cpp is not installed."
log_info "Install from AUR for automatic per-process priority tuning:"
log_info " yay -S ananicy-cpp cachyos-ananicy-rules-git"
return 0
}
# ===================================================================
# Apply all tweaks
# ===================================================================
main() {
apply_tweak \
"Tweak 1/10: Set CPU governor to performance on all cores" \
tweak_cpu_governor
apply_tweak \
"Tweak 2/10: Optimize I/O scheduler per drive type" \
tweak_io_scheduler
apply_tweak \
"Tweak 3/10: Tune VM/memory sysctl for desktop responsiveness" \
tweak_vm_sysctl
apply_tweak \
"Tweak 4/10: Enable TCP BBR + fastopen + larger buffers" \
tweak_network_sysctl
apply_tweak \
"Tweak 5/10: Enable fstrim.timer for SSD TRIM maintenance" \
tweak_fstrim
apply_tweak \
"Tweak 6/10: NVIDIA GPU persistence mode + max performance" \
tweak_nvidia_gpu
apply_tweak \
"Tweak 7/10: CPU vulnerability mitigations (--aggressive only)" \
tweak_mitigations
apply_tweak \
"Tweak 8/10: Disable NetworkManager-wait-online (faster boot)" \
tweak_nm_wait_online
apply_tweak \
"Tweak 9/10: Vacuum & cap systemd journal at 300M" \
tweak_journal
apply_tweak \
"Tweak 10/10: Enable ananicy-cpp process prioritization" \
tweak_ananicy
# ---------------------------------------------------------------
# Summary
# ---------------------------------------------------------------
echo ""
echo "=============================="
echo " Desktop Optimization Summary"
echo "=============================="
if [[ $DRY_RUN == "true" ]]; then
log_info "Dry-run mode — no changes were made."
else
log_ok "Optimizations applied: $TWEAKS_APPLIED"
if [[ $TWEAKS_SKIPPED -gt 0 ]]; then
log_warn "Optimizations skipped: $TWEAKS_SKIPPED"
fi
fi
echo ""
# Advisory: check for noatime
local root_mount_opts
root_mount_opts=$(findmnt -n -o OPTIONS / 2>/dev/null || true)
if [[ -n $root_mount_opts ]] && ! echo "$root_mount_opts" | grep -q 'noatime'; then
log_info "Tip: Your root filesystem does not use 'noatime'."
log_info " Adding 'noatime' to /etc/fstab can reduce unnecessary disk writes."
log_info " (Change 'relatime' or 'atime' to 'noatime' in /etc/fstab, then reboot)"
fi
if [[ $AGGRESSIVE == "true" ]]; then
log_warn "Aggressive mode was used — mitigations=off trades security for speed."
fi
echo ""
log_info "Reboot recommended for kernel parameter and boot loader changes to take effect."
log_info "Verify after reboot with: diagnose_arch_performance.sh"
}
main