wangbo/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2026-04-15 13:02:35 +08:00
Code Issues Actions 17 Packages Projects Releases Wiki Activity

feat(entrypoint): GPU-first single log; UV-free; CM_* driven Manager config; standardized SageAttention logs

Browse Source 
Move GPU probe to the top with a single comprehensive device report (index, name, CC, VRAM) and early exit if no compatible GPUs (CC >= 7.5 not met); avoid duplicate logs after user switch via an internal flag. Remove uv usage and ensurepip, keeping system-wide pip installs only. Add CM_* environment variable handling to seed and reconcile ComfyUI-Manager’s persistent config.ini under user/default/ComfyUI-Manager on first and subsequent boots. Standardize “SageAttention” naming and reduce duplicate “building” messages; keep runtime Triton adjustment only when needed for Turing.
This commit is contained in:
clsferguson 2025-10-02 20:47:07 -06:00 committed by GitHub
parent d1ba18dac8
commit de6351e9bf
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 131 additions and 74 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

205
entrypoint.sh
View File

@ -12,6 +12,9 @@ SAGE_ATTENTION_DIR="$BASE_DIR/.sage_attention"
SAGE_ATTENTION_BUILT_FLAG="$SAGE_ATTENTION_DIR/.built"
PERMISSIONS_SET_FLAG="$BASE_DIR/.permissions_set"
FIRST_RUN_FLAG="$BASE_DIR/.first_run_done"
CFG_DIR="$BASE_DIR/user/default/ComfyUI-Manager"
CFG_FILE="$CFG_DIR/config.ini"
CFG_SEEDED_FLAG="$CFG_DIR/.cm_seeded"
# --- logging ---
log() { echo "[$(date '+%H:%M:%S')] $1"; }
@ -51,12 +54,21 @@ if not torch.cuda.is_available():
print("SAGE_STRATEGY='fallback'"); sys.exit(0)
n = torch.cuda.device_count()
ccs = []
names = []
vram = []
flags = {"DET_TURING":False,"DET_AMP80":False,"DET_AMP86":False,"DET_AMP87":False,"DET_ADA":False,"DET_HOPPER":False,"DET_BW12":False,"DET_BW10":False}
compat = False
for i in range(n):
p = torch.cuda.get_device_properties(i)
mj, mn = p.major, p.minor
ccs.append(f"{mj}.{mn}")
c = f"{mj}.{mn}"
ccs.append(c)
names.append(p.name)
try:
mem_mib = int(p.total_memory // (1024**2))
except Exception:
mem_mib = 0
vram.append(str(mem_mib))
if (mj,mn)==(7,5): flags["DET_TURING"]=True
elif (mj,mn)==(8,0): flags["DET_AMP80"]=True
elif (mj,mn)==(8,6): flags["DET_AMP86"]=True
@ -88,7 +100,13 @@ print(f"TORCH_CUDA_ARCH_LIST='{arch_list}'")
for k,v in flags.items():
print(f"{k}={'true' if v else 'false'}")
print(f"SAGE_STRATEGY='{strategy}'")
print(f"[GPU] Found {n} CUDA device(s); CC list: {arch_list or 'none'}; strategy={strategy}; compat>=7.5:{compat}", file=sys.stderr)
# Detailed device report to stderr (single, comprehensive log)
if n == 0:
print(f\"[GPU] Found 0 CUDA device(s)\", file=sys.stderr)
else:
print(f\"[GPU] Found {n} CUDA device(s); CC list: {arch_list or 'none'}; strategy={strategy}; compat>=7.5:{compat}\", file=sys.stderr)
for i,(nm,cc,vm) in enumerate(zip(names, ccs, vram)):
print(f\"[GPU] #{i}: {nm} | CC {cc} | VRAM {vm} MiB\", file=sys.stderr)
PY
}
@ -107,45 +125,73 @@ PY
case "${SAGE_STRATEGY:-fallback}" in
"mixed_with_turing"|"turing_only")
if [ "$cur" != "3.2.0" ]; then
log "Installing Triton 3.2.0 for Turing compatibility (current: ${cur:-none})"
log "Setting Triton to 3.2.0 for Turing compatibility (current: ${cur:-none})"
python -m pip install --no-cache-dir "triton==3.2.0" || true
else
log "Triton 3.2.0 already present; skipping"
fi
;;
*)
# Image bakes Triton==3.4.0; leave as-is
# Image bakes Triton==3.4.0; leave as-is for Ampere/Ada/Hopper/Blackwell
log "Using baked Triton (${cur:-unknown}); no change"
;;
esac
}
build_sage_attention_mixed() {
log "Building Sage Attention..."
mkdir -p "$SAGE_ATTENTION_DIR"; cd "$SAGE_ATTENTION_DIR"
test_sage_attention() {
python - <<'PY' 2>/dev/null
import sys
try:
import sageattention
print("[TEST] SageAttention import: SUCCESS")
v=getattr(sageattention,'__version__',None)
if v: print(f"[TEST] Version: {v}")
sys.exit(0)
except ImportError as e:
print(f"[TEST] SageAttention import: FAILED - {e}")
sys.exit(1)
except Exception as e:
print(f"[TEST] SageAttention test: ERROR - {e}")
sys.exit(1)
PY
}
build_sage_attention() {
mkdir -p "$SAGE_ATTENTION_DIR"
cd "$SAGE_ATTENTION_DIR"
export TORCH_CUDA_ARCH_LIST="${SAGE_ARCH_LIST_OVERRIDE:-${TORCH_CUDA_ARCH_LIST:-}}"
if [ -z "${TORCH_CUDA_ARCH_LIST:-}" ]; then
TORCH_CUDA_ARCH_LIST="8.0;8.6;8.9;9.0;10.0;12.0"
fi
log "Set TORCH_CUDA_ARCH_LIST=$TORCH_CUDA_ARCH_LIST"
log "Building SageAttention (ARCHS=$TORCH_CUDA_ARCH_LIST)"
case "${SAGE_STRATEGY:-fallback}" in
"mixed_with_turing"|"turing_only")
log "Cloning SageAttention v1.0 for Turing"
if [ -d "SageAttention/.git" ]; then cd SageAttention; git fetch --depth 1 origin || return 1; git checkout v1.0 2>/dev/null || git checkout -b v1.0 origin/v1.0 || return 1; git reset --hard origin/v1.0 || return 1
else rm -rf SageAttention; git clone --depth 1 https://github.com/thu-ml/SageAttention.git -b v1.0 || return 1; cd SageAttention; fi
if [ -d "SageAttention/.git" ]; then
cd SageAttention
git fetch --depth 1 origin || return 1
git checkout v1.0 2>/dev/null || git checkout -b v1.0 origin/v1.0 || return 1
git reset --hard origin/v1.0 || return 1
else
rm -rf SageAttention
git clone --depth 1 https://github.com/thu-ml/SageAttention.git -b v1.0 || return 1
cd SageAttention
fi
;;
*)
log "Cloning latest SageAttention"
if [ -d "SageAttention/.git" ]; then cd SageAttention; git fetch --depth 1 origin || return 1; git reset --hard origin/main || return 1
else rm -rf SageAttention; git clone --depth 1 https://github.com/thu-ml/SageAttention.git || return 1; cd SageAttention; fi
if [ -d "SageAttention/.git" ]; then
cd SageAttention
git fetch --depth 1 origin || return 1
git reset --hard origin/main || return 1
else
rm -rf SageAttention
git clone --depth 1 https://github.com/thu-ml/SageAttention.git || return 1
cd SageAttention
fi
;;
esac
[ "${SAGE_VERBOSE_BUILD:-0}" = "1" ] && export TORCH_CPP_BUILD_VERBOSE=1
local jobs; jobs="$(decide_build_jobs)"
log "Using MAX_JOBS=${jobs} for SageAttention build"
log "Compiling SageAttention (MAX_JOBS=${jobs})"
if MAX_JOBS="${jobs}" python -m pip install --no-build-isolation .; then
echo "${SAGE_STRATEGY:-fallback}|${TORCH_CUDA_ARCH_LIST:-}" > "$SAGE_ATTENTION_BUILT_FLAG"
log "SageAttention built successfully"
@ -164,49 +210,79 @@ needs_rebuild() {
return 1
}
test_sage_attention() {
python -c "
import sys
try:
import sageattention; print('[TEST] SageAttention import: SUCCESS')
v=getattr(sageattention,'__version__',None)
if v: print(f'[TEST] Version: {v}'); sys.exit(0)
except ImportError as e:
print(f'[TEST] SageAttention import: FAILED - {e}'); sys.exit(1)
except Exception as e:
print(f'[TEST] SageAttention test: ERROR - {e}'); sys.exit(1)
" 2>/dev/null
}
setup_sage_attention() {
export SAGE_ATTENTION_BUILT=0 SAGE_ATTENTION_AVAILABLE=0
if [ "${GPU_COUNT:-0}" -eq 0 ]; then log "No GPUs detected, skipping SageAttention setup"; return 0; fi
if [ "${GPU_COUNT:-0}" -eq 0 ]; then log "No GPUs detected, skipping SageAttention"; return 0; fi
if [ "${COMPAT_GE_75:-0}" -ne 1 ]; then log "GPU compute capability < 7.5; skipping SageAttention"; return 0; fi
if needs_rebuild || ! test_sage_attention; then
log "Building SageAttention..."
if install_triton_version && build_sage_attention_mixed && test_sage_attention; then
if needs_rebuild || ! test_sage_attention >/dev/null; then
log "Preparing SageAttention"
if install_triton_version && build_sage_attention && test_sage_attention >/dev/null; then
export SAGE_ATTENTION_BUILT=1 SAGE_ATTENTION_AVAILABLE=1
log "SageAttention is built; set FORCE_SAGE_ATTENTION=1 to enable it at startup"
log "SageAttention is available; set FORCE_SAGE_ATTENTION=1 to enable at startup"
else
export SAGE_ATTENTION_BUILT=0 SAGE_ATTENTION_AVAILABLE=0
log "WARNING: SageAttention is not available after build attempt"
fi
else
export SAGE_ATTENTION_BUILT=1 SAGE_ATTENTION_AVAILABLE=1
log "SageAttention already built and importable"
log "SageAttention already available"
fi
# Strategy is only needed for build decisions; clear after setup to avoid confusion
unset SAGE_STRATEGY
}
# --- Manager config from CM_* env ---
configure_manager_from_env() {
mkdir -p "$CFG_DIR" || true
# Collect CM_* into an INI under [default]
# First-boot: replace config.ini; subsequent: reconcile keys differing from env
python - "$CFG_FILE" "$CFG_SEEDED_FLAG" <<'PY'
import os, sys, configparser, pathlib
cfg_file = pathlib.Path(sys.argv[1])
seed_flag = pathlib.Path(sys.argv[2])
# Collect CM_* environment variables
env_pairs = {}
for k,v in os.environ.items():
if not k.startswith("CM_"): continue
key = k[3:].lower()
env_pairs[key] = v
cfg = configparser.ConfigParser()
if not seed_flag.exists() or not cfg_file.exists():
# First-boot: (re)create from CM_* exclusively
cfg["default"] = {}
for k,v in env_pairs.items():
cfg["default"][k] = v
cfg_file.write_text("", encoding="utf-8")
with cfg_file.open("w", encoding="utf-8") as f:
cfg.write(f)
seed_flag.parent.mkdir(parents=True, exist_ok=True)
seed_flag.write_text("seeded", encoding="utf-8")
sys.exit(0)
# Subsequent boots: reconcile only CM_* keys
cfg.read(cfg_file, encoding="utf-8")
if "default" not in cfg: cfg["default"] = {}
changed = False
for k,v in env_pairs.items():
if cfg["default"].get(k) != v:
cfg["default"][k] = v
changed = True
if changed:
with cfg_file.open("w", encoding="utf-8") as f:
cfg.write(f)
PY
}
# --- early GPU probe and exit (before heavy setup) ---
eval "$(probe_and_prepare_gpu)"
log "GPU probe: ${GPU_COUNT:-0} CUDA device(s); CC list: ${TORCH_CUDA_ARCH_LIST:-none}; strategy=${SAGE_STRATEGY:-fallback}"
if [ "${GPU_COUNT:-0}" -eq 0 ]; then
log "No NVIDIA GPU detected; shutting down."
exit 0
fi
if [ "${COMPAT_GE_75:-0}" -ne 1 ]; then
log "GPU compute capability < 7.5; shutting down."
exit 0
if [ -z "${SKIP_EARLY_PROBE:-}" ]; then
eval "$(probe_and_prepare_gpu)"
if [ "${GPU_COUNT:-0}" -eq 0 ]; then
log "No NVIDIA GPU detected; shutting down."
exit 0
fi
if [ "${COMPAT_GE_75:-0}" -ne 1 ]; then
log "GPU compute capability < 7.5; shutting down."
exit 0
fi
fi
# --- root to runtime user ---
@ -219,8 +295,7 @@ if [ "$(id -u)" = "0" ]; then
usermod -o -u "${PUID}" "${APP_USER}" || true
mkdir -p "/home/${APP_USER}"
for d in "$BASE_DIR" "/home/$APP_USER"; do [ -e "$d" ] && chown -R "${APP_USER}:${APP_GROUP}" "$d" || true; done
# Make system site-packages writable by the runtime user (no venvs; system-wide installs)
# Make system site-packages writable by the runtime user (system-wide installs; no venvs)
readarray -t PY_PATHS < <(python - <<'PY'
import sys, sysconfig, os, site, datetime
def log(m): print(f"[bootstrap:python {datetime.datetime.now().strftime('%H:%M:%S')}] {m}", file=sys.stderr, flush=True)
@ -250,44 +325,26 @@ PY
chown -R "${APP_USER}:${APP_GROUP}" "$d" || true
chmod -R u+rwX,g+rwX "$d" || true
done
if [ -d "/usr/local/lib/python3.12/site-packages" ]; then
chown -R "${APP_USER}:${APP_GROUP}" /usr/local/lib/python3.12/site-packages || true
chmod -R u+rwX,g+rwX /usr/local/lib/python3.12/site-packages || true
fi
touch "$PERMISSIONS_SET_FLAG"; chown "${APP_USER}:${APP_GROUP}" "$PERMISSIONS_SET_FLAG"
log "User permissions configured"
else
log "User permissions already configured, skipping..."
fi
exec runuser -u "${APP_USER}" -- "$0" "$@"
exec env SKIP_EARLY_PROBE=1 runuser -u "${APP_USER}" -- "$0" "$@"
fi
# From here on, running as $APP_USER
export PATH="$HOME/.local/bin:$PATH"
pyver="$(python -c 'import sys; print(f"{sys.version_info.major}.{sys.version_info.minor}")')"
export PYTHONPATH="$HOME/.local/lib/python${pyver}/site-packages:${PYTHONPATH:-}"
export PIP_PREFER_BINARY=1
# --- refresh GPU probe after user switch (no exit) ---
# Probe again silently to set build variables (no duplicate GPU logs)
eval "$(probe_and_prepare_gpu)"
log "GPU probe (post-switch): ${GPU_COUNT:-0} CUDA device(s); CC list: ${TORCH_CUDA_ARCH_LIST:-none}; strategy=${SAGE_STRATEGY:-fallback}"
# Ensure pip works
python -m pip --version >/dev/null 2>&1 || python -m ensurepip --upgrade >/dev/null 2>&1 || true
python -m pip --version >/dev/null 2>&1 || log "WARNING: pip still not available after ensurepip"
# Ensure minimal Python deps for ComfyUI-Manager (pre-baked, but verify)
python - <<'PY' || python -m pip install --no-cache-dir toml GitPython || true
import sys
import importlib
for m in ("toml","git"):
try: importlib.import_module(m)
except Exception: sys.exit(1)
sys.exit(0)
PY
# Configure ComfyUI-Manager from CM_* env
configure_manager_from_env
# --- SageAttention setup using probed data ---
setup_sage_attention
@ -306,7 +363,7 @@ fi
# --- first-run install of custom_nodes ---
if [ ! -f "$FIRST_RUN_FLAG" ] || [ "${COMFY_FORCE_INSTALL:-0}" = "1" ]; then
if [ "${COMFY_AUTO_INSTALL:-1}" = "1" ]; then
log "First run or forced; installing custom node dependencies..."
log "Installing custom node dependencies (first run/forced)"
shopt -s nullglob
for d in "$CUSTOM_NODES_DIR"/*; do
[ -d "$d" ] || continue
@ -333,12 +390,12 @@ fi
# --- launch ComfyUI ---
COMFYUI_ARGS=""
if [ "${FORCE_SAGE_ATTENTION:-0}" = "1" ] && test_sage_attention; then
if [ "${FORCE_SAGE_ATTENTION:-0}" = "1" ] && test_sage_attention >/dev/null; then
COMFYUI_ARGS="--use-sage-attention"
log "Starting ComfyUI with SageAttention (FORCE_SAGE_ATTENTION=1)"
else
if [ "${SAGE_ATTENTION_AVAILABLE:-0}" = "1" ]; then
log "SageAttention is built; set FORCE_SAGE_ATTENTION=1 to enable"
log "SageAttention is available; set FORCE_SAGE_ATTENTION=1 to enable"
else
log "SageAttention not available; starting without it"
fi