diff options
Diffstat (limited to 'tools/guest-diag/run_traced.sh')
| -rwxr-xr-x | tools/guest-diag/run_traced.sh | 53 |
1 files changed, 53 insertions, 0 deletions
diff --git a/tools/guest-diag/run_traced.sh b/tools/guest-diag/run_traced.sh new file mode 100755 index 0000000..e8d3a09 --- /dev/null +++ b/tools/guest-diag/run_traced.sh @@ -0,0 +1,53 @@ +#!/bin/bash +# run_traced.sh - run cdt under the ioctl tracer, keep the libcuda tracker page +# mapped (pagemap + sema_map_add) and poke UVM's tracking semaphore forward as +# the sweep completes pushes, so the run reaches the rc=999 rollback where the +# tracer records the failing driver call. +set -u +cd "$(dirname "$(readlink -f "$0")")/../.." + +EXEC() { timeout 30 ./scripts/vgpu_vm.py exec --timeout 25 "$1" 2>/dev/null; } + +EXEC 'sudo pkill -9 -x cdt; sudo pkill -9 -x ptio; sleep 1; sudo rm -f /tmp/cdt.out /tmp/tr.log; sudo setsid /tmp/ptio /tmp/cdt > /tmp/cdt.out 2> /tmp/tr.log & sleep 3; pgrep -x cdt | head -1' + +for i in $(seq 1 40); do + sleep 10 + PID=$(EXEC 'pgrep -x cdt | head -1' | tail -1) + [ -n "$PID" ] || { echo "[*] cdt gone at $((i*10))s"; break; } + + # pin the tracker page when the guest has it mapped + PA=$(EXEC "sudo python3 -c \" +import os +fd=os.open('/proc/$PID/pagemap', os.O_RDONLY) +os.lseek(fd, (0x20480f000//4096)*8, 0) +e=int.from_bytes(os.read(fd,8),'little') +os.close(fd) +pfn=e & ((1<<55)-1) +print('PA=0x%x'%((pfn<<12)|0xeb0)) +\"" | grep -o 'PA=0x[0-9a-f]*' | head -1 | sed 's/PA=//') + if [ -n "$PA" ]; then + sudo sh -c "echo 0x20480feb0,$PA > /sys/module/open_iov_vfio/parameters/sema_map_add" 2>/dev/null + fi + + # nudge UVM's tracking semaphore forward when the sweep has stalled + sudo sh -c 'echo 0x121000010,1 > /sys/module/open_iov_vfio/parameters/sema_probe' 2>/dev/null + sleep 1 + VALS=$(sudo dmesg -T | grep -A4 "sema-probe: va=0x121000010" | tail -4) + SPA=$(echo "$VALS" | grep -oE "pa=0x[0-9a-f]+" | head -1 | cut -d= -f2) + CUR=$(echo "$VALS" | grep -oE "\[0\] 0x[0-9a-f]+" | head -1 | awk '{print $2}') + if [ -n "$SPA" ] && [ -n "$CUR" ]; then + NEXT=$(printf '0x%x' $(( CUR + 1 ))) + sudo sh -c "echo 0x$SPA,$NEXT > /sys/module/open_iov_vfio/parameters/poke_guest" 2>/dev/null + echo " [$((i*10))s] cdt=$PID sema $CUR -> $NEXT" + else + echo " [$((i*10))s] cdt=$PID (sema unresolved)" + fi + + STATE=$(EXEC 'tail -1 /tmp/cdt.out' | tail -1) + case "$STATE" in + *FAILED*|*"ALL OK"*|*cuMemAlloc*) echo "[*] finished: $STATE"; break ;; + esac +done + +echo "=== trace tail ===" +EXEC 'tail -25 /tmp/tr.log' |
