From cf74193cc37f2ace1197b1e6ebacc6fe542767c8 Mon Sep 17 00:00:00 2001

From: Kairui Song <kasong@redhat.com>

Date: Wed, 8 Apr 2020 16:39:52 +0800

Subject: [PATCH] Remove memtrace-ko and rd.memdebug=4 support in dracut

This feature could be off loaded to memstrack, which have better

accurecy, better performance, and have more detailed tracing features.

Also simplify make_trace_mem a bit.

And currently rd.memdebug=4 is unstable, fails from time to time.

(cherry picked from commit 49c4172f4eef6e2015615e132b199a7ec0699ffc)

Resolves: #1829528

---

 dracut.cmdline.7.asc                             |   4 +-

 modules.d/98dracut-systemd/dracut-cmdline.sh     |   2 +-

 modules.d/98dracut-systemd/dracut-pre-mount.sh   |   2 +-

 modules.d/98dracut-systemd/dracut-pre-pivot.sh   |   2 +-

 modules.d/98dracut-systemd/dracut-pre-trigger.sh |   2 +-

 modules.d/99base/dracut-lib.sh                   |  44 +-----

 modules.d/99base/init.sh                         |   8 +-

 modules.d/99base/memtrace-ko.sh                  | 191 -----------------------

 modules.d/99base/module-setup.sh                 |   1 -

 9 files changed, 18 insertions(+), 238 deletions(-)

diff --git a/dracut.cmdline.7.asc b/dracut.cmdline.7.asc

index ab9a24ad..0dc58d1a 100644

--- a/dracut.cmdline.7.asc

+++ b/dracut.cmdline.7.asc

@@ -187,9 +187,9 @@ It should be attached to any report about dracut problems.

     _/run/initramfs/init.log_.

     If "quiet" is set, it also logs to the console.

-**rd.memdebug=[0-4]**::

+**rd.memdebug=[0-3]**::

     Print memory usage info at various points, set the verbose level from 0 to 4.

-+    

++

     Higher level means more debugging output:

 +

 ----

diff --git a/modules.d/98dracut-systemd/dracut-cmdline.sh b/modules.d/98dracut-systemd/dracut-cmdline.sh

index bff9435a..6c6ee026 100755

--- a/modules.d/98dracut-systemd/dracut-cmdline.sh

+++ b/modules.d/98dracut-systemd/dracut-cmdline.sh

@@ -42,7 +42,7 @@ export root

 export rflags

 export fstype

-make_trace_mem "hook cmdline" '1+:mem' '1+:iomem' '3+:slab' '4+:komem'

+make_trace_mem "hook cmdline" '1+:mem' '1+:iomem' '3+:slab'

 # run scriptlets to parse the command line

 getarg 'rd.break=cmdline' -d 'rdbreak=cmdline' && emergency_shell -n cmdline "Break before cmdline"

 source_hook cmdline

diff --git a/modules.d/98dracut-systemd/dracut-pre-mount.sh b/modules.d/98dracut-systemd/dracut-pre-mount.sh

index a3b9d291..ae511286 100755

--- a/modules.d/98dracut-systemd/dracut-pre-mount.sh

+++ b/modules.d/98dracut-systemd/dracut-pre-mount.sh

@@ -8,7 +8,7 @@ type getarg >/dev/null 2>&1 || . /lib/dracut-lib.sh

 source_conf /etc/conf.d

-make_trace_mem "hook pre-mount" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-mount" '1:shortmem' '2+:mem' '3+:slab'

 # pre pivot scripts are sourced just before we doing cleanup and switch over

 # to the new root.

 getarg 'rd.break=pre-mount' 'rdbreak=pre-mount' && emergency_shell -n pre-mount "Break pre-mount"

diff --git a/modules.d/98dracut-systemd/dracut-pre-pivot.sh b/modules.d/98dracut-systemd/dracut-pre-pivot.sh

index dc9a2504..96e6f8ca 100755

--- a/modules.d/98dracut-systemd/dracut-pre-pivot.sh

+++ b/modules.d/98dracut-systemd/dracut-pre-pivot.sh

@@ -8,7 +8,7 @@ type getarg >/dev/null 2>&1 || . /lib/dracut-lib.sh

 source_conf /etc/conf.d

-make_trace_mem "hook pre-pivot" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-pivot" '1:shortmem' '2+:mem' '3+:slab'

 # pre pivot scripts are sourced just before we doing cleanup and switch over

 # to the new root.

 getarg 'rd.break=pre-pivot' 'rdbreak=pre-pivot' && emergency_shell -n pre-pivot "Break pre-pivot"

diff --git a/modules.d/98dracut-systemd/dracut-pre-trigger.sh b/modules.d/98dracut-systemd/dracut-pre-trigger.sh

index 7cd821ed..a1a33960 100755

--- a/modules.d/98dracut-systemd/dracut-pre-trigger.sh

+++ b/modules.d/98dracut-systemd/dracut-pre-trigger.sh

@@ -8,7 +8,7 @@ type getarg >/dev/null 2>&1 || . /lib/dracut-lib.sh

 source_conf /etc/conf.d

-make_trace_mem "hook pre-trigger" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-trigger" '1:shortmem' '2+:mem' '3+:slab'

 source_hook pre-trigger

diff --git a/modules.d/99base/dracut-lib.sh b/modules.d/99base/dracut-lib.sh

index f71e757c..502c7568 100755

--- a/modules.d/99base/dracut-lib.sh

+++ b/modules.d/99base/dracut-lib.sh

@@ -1220,50 +1220,25 @@ are_lists_eq() {

 setmemdebug() {

     if [ -z "$DEBUG_MEM_LEVEL" ]; then

-        export DEBUG_MEM_LEVEL=$(getargnum 0 0 4 rd.memdebug)

+        export DEBUG_MEM_LEVEL=$(getargnum 0 0 3 rd.memdebug)

     fi

 }

 setmemdebug

-cleanup_trace_mem()

-{

-    # tracekomem based on kernel trace needs cleanup after use.

-    if [ "$DEBUG_MEM_LEVEL" -eq 4 ]; then

-        tracekomem --cleanup

-    fi

-}

-

-# parameters: msg [trace_level:trace]...

-make_trace_mem()

-{

-    local msg

-    msg="$1"

-    shift

-    if [ -n "$DEBUG_MEM_LEVEL" ] && [ "$DEBUG_MEM_LEVEL" -gt 0 ]; then

-        make_trace show_memstats $DEBUG_MEM_LEVEL "[debug_mem]" "$msg" "$@" >&2

-    fi

-}

-

 # parameters: func log_level prefix msg [trace_level:trace]...

-make_trace()

+make_trace_mem()

 {

-    local func log_level prefix msg msg_printed

+    local log_level prefix msg msg_printed

     local trace trace_level trace_in_higher_levels insert_trace

-    func=$1

-    shift

-

-    log_level=$1

-    shift

-

-    prefix=$1

-    shift

-

     msg=$1

     shift

-    if [ -z "$log_level" ]; then

+    prefix='[debug_mem]'

+    log_level=$DEBUG_MEM_LEVEL

+

+    if [ -z "$log_level" ] || [ "$log_level" -le 0 ]; then

         return

     fi

@@ -1296,7 +1271,7 @@ make_trace()

                 echo "$prefix $msg"

                 msg_printed=1

             fi

-            $func $trace

+            show_memstats $trace

         fi

         shift

     done

@@ -1318,9 +1293,6 @@ show_memstats()

         iomem)

             cat /proc/iomem

             ;;

-        komem)

-            tracekomem

-            ;;

     esac

 }

diff --git a/modules.d/99base/init.sh b/modules.d/99base/init.sh

index e4f7cff1..2c0ccd66 100755

--- a/modules.d/99base/init.sh

+++ b/modules.d/99base/init.sh

@@ -131,7 +131,7 @@ if ! getargbool 1 'rd.hostonly'; then

 fi

 # run scriptlets to parse the command line

-make_trace_mem "hook cmdline" '1+:mem' '1+:iomem' '3+:slab' '4+:komem'

+make_trace_mem "hook cmdline" '1+:mem' '1+:iomem' '3+:slab'

 getarg 'rd.break=cmdline' -d 'rdbreak=cmdline' && emergency_shell -n cmdline "Break before cmdline"

 source_hook cmdline

@@ -160,7 +160,7 @@ fi

 udevproperty "hookdir=$hookdir"

-make_trace_mem "hook pre-trigger" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-trigger" '1:shortmem' '2+:mem' '3+:slab'

 getarg 'rd.break=pre-trigger' -d 'rdbreak=pre-trigger' && emergency_shell -n pre-trigger "Break before pre-trigger"

 source_hook pre-trigger

@@ -230,7 +230,7 @@ unset RDRETRY

 # pre-mount happens before we try to mount the root filesystem,

 # and happens once.

-make_trace_mem "hook pre-mount" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-mount" '1:shortmem' '2+:mem' '3+:slab'

 getarg 'rd.break=pre-mount' -d 'rdbreak=pre-mount' && emergency_shell -n pre-mount "Break pre-mount"

 source_hook pre-mount

@@ -266,7 +266,7 @@ done

 # pre pivot scripts are sourced just before we doing cleanup and switch over

 # to the new root.

-make_trace_mem "hook pre-pivot" '1:shortmem' '2+:mem' '3+:slab' '4+:komem'

+make_trace_mem "hook pre-pivot" '1:shortmem' '2+:mem' '3+:slab'

 getarg 'rd.break=pre-pivot' -d 'rdbreak=pre-pivot' && emergency_shell -n pre-pivot "Break pre-pivot"

 source_hook pre-pivot

diff --git a/modules.d/99base/memtrace-ko.sh b/modules.d/99base/memtrace-ko.sh

deleted file mode 100755

index ee035e15..00000000

--- a/modules.d/99base/memtrace-ko.sh

+++ /dev/null

@@ -1,191 +0,0 @@

-#!/bin/sh

-

-# Try to find out kernel modules with large total memory allocation during loading.

-# For large slab allocation, it will fall into buddy, also not trace "mm_page_free"

-# considering large free is quite rare for module_init, thus saving tons of events

-# to avoid trace data overwritten.

-#

-# Therefore, tracing "mm_page_alloc"alone should be enough for the purpose.

-

-# "sys/kernel/tracing" has the priority if exists.

-get_trace_base() {

-    # trace access through debugfs would be obsolete if "/sys/kernel/tracing" is available.

-    if [ -d "/sys/kernel/tracing" ]; then

-        echo "/sys/kernel"

-    else

-        echo "/sys/kernel/debug"

-    fi

-}

-

-# We want to enable these trace events.

-get_want_events() {

-    echo "module:module_put module:module_load kmem:mm_page_alloc"

-}

-

-get_event_filter() {

-    echo "comm == systemd-udevd || comm == modprobe || comm == insmod"

-}

-

-is_trace_ready() {

-    local trace_base want_events current_events

-

-    trace_base=$(get_trace_base)

-    ! [ -f "$trace_base/tracing/trace" ] && return 1

-

-    [ "$(cat $trace_base/tracing/tracing_on)" -eq 0 ] && return 1

-

-    # Also check if trace events were properly setup.

-    want_events=$(get_want_events)

-    current_events=$(echo $(cat $trace_base/tracing/set_event))

-    [ "$current_events" != "$want_events" ] && return 1

-

-    return 0

-}

-

-prepare_trace() {

-    local trace_base

-

-    trace_base=$(get_trace_base)

-    # old debugfs interface case.

-    if ! [ -d "$trace_base/tracing" ]; then

-        mount none -t debugfs $trace_base

-    # new tracefs interface case.

-    elif ! [ -f "$trace_base/tracing/trace" ]; then

-        mount none -t tracefs "$trace_base/tracing"

-    fi

-

-    if ! [ -f "$trace_base/tracing/trace" ]; then

-        echo "WARN: Mount trace failed for kernel module memory analyzing."

-        return 1

-    fi

-

-    # Active all the wanted trace events.

-    echo "$(get_want_events)" > $trace_base/tracing/set_event

-

-    # There are three kinds of known applications for module loading:

-    # "systemd-udevd", "modprobe" and "insmod".

-    # Set them as the global events filter.

-    # NOTE: Some kernel may not support this format of filter, anyway

-    #       the operation will fail and it doesn't matter.

-    echo "$(get_event_filter)" > $trace_base/tracing/events/kmem/filter 2>&1

-    echo "$(get_event_filter)" > $trace_base/tracing/events/module/filter 2>&1

-

-    # Set the number of comm-pid if supported.

-    if [ -f "$trace_base/tracing/saved_cmdlines_size" ]; then

-        # Thanks to filters, 4096 is big enough(also well supported).

-        echo 4096 > $trace_base/tracing/saved_cmdlines_size

-    fi

-

-    # Enable and clear trace data for the first time.

-    echo 1 > $trace_base/tracing/tracing_on

-    echo > $trace_base/tracing/trace

-    echo "Prepare trace success."

-    return 0

-}

-

-order_to_pages()

-{

-    local pages=1

-    local order=$1

-

-    while [ "$order" != 0 ]; do

-        order=$((order-1))

-        pages=$(($pages*2))

-	done

-

-    echo $pages

-}

-

-parse_trace_data() {

-    local module_name tmp_eval pages

-

-    cat "$(get_trace_base)/tracing/trace" | while read pid cpu flags ts function args

-    do

-        # Skip comment lines

-        if [ "$pid" = "#" ]; then

-            continue

-        fi

-

-        pid=${pid##*-}

-        function=${function%:}

-        if [ "$function" = "module_load" ]; then

-            # One module is being loaded, save the task pid for tracking.

-            # Remove the trailing after whitespace, there may be the module flags.

-            module_name=${args%% *}

-            # Mark current_module to track the task.

-            eval current_module_$pid="$module_name"

-            tmp_eval=$(eval echo '${module_loaded_'${module_name}'}')

-            if [ -n "$tmp_eval" ]; then

-                echo "WARN: \"$module_name\" was loaded multiple times!"

-            fi

-            eval unset module_loaded_$module_name

-            eval nr_alloc_pages_$module_name=0

-            continue

-        fi

-

-        module_name=$(eval echo '${current_module_'${pid}'}')

-        if [ -z "$module_name" ]; then

-            continue

-        fi

-

-        # Once we get here, the task is being tracked(is loading a module).

-        if [ "$function" = "module_put" ]; then

-            # Mark the module as loaded when the first module_put event happens after module_load.

-            tmp_eval=$(eval echo '${nr_alloc_pages_'${module_name}'}')

-            echo "$tmp_eval pages consumed by \"$module_name\""

-            eval module_loaded_$module_name=1

-            # Module loading finished, so untrack the task.

-            eval unset current_module_$pid

-            eval unset nr_alloc_pages_$module_name

-            continue

-        fi

-

-        if [ "$function" = "mm_page_alloc" ]; then

-            # Get order first, then convert to actual pages.

-            pages=$(echo $args | sed -e 's/.*order=\([0-9]*\) .*/\1/')

-            pages=$(order_to_pages "$pages")

-            tmp_eval=$(eval echo '${nr_alloc_pages_'${module_name}'}')

-            eval nr_alloc_pages_$module_name="$(($tmp_eval+$pages))"

-        fi

-    done

-}

-

-cleanup_trace() {

-    local trace_base

-

-    if is_trace_ready; then

-        trace_base=$(get_trace_base)

-        echo 0 > $trace_base/tracing/tracing_on

-        echo > $trace_base/tracing/trace

-        echo > $trace_base/tracing/set_event

-        echo 0 > $trace_base/tracing/events/kmem/filter

-        echo 0 > $trace_base/tracing/events/module/filter

-    fi

-}

-

-show_usage() {

-    echo "Find out kernel modules with large memory consumption during loading based on trace."

-    echo "Usage:"

-    echo "1) run it first to setup trace."

-    echo "2) run again to parse the trace data if any."

-    echo "3) run with \"--cleanup\" option to cleanup trace after use."

-}

-

-if [ "$1" = "--help" ]; then

-    show_usage

-    exit 0

-fi

-

-if [ "$1" = "--cleanup" ]; then

-    cleanup_trace

-    exit 0

-fi

-

-if is_trace_ready ; then

-    echo "tracekomem - Rough memory consumption by loading kernel modules (larger value with better accuracy)"

-    parse_trace_data

-else

-    prepare_trace

-fi

-

-exit $?

diff --git a/modules.d/99base/module-setup.sh b/modules.d/99base/module-setup.sh

index c9ead01d..3578643c 100755

--- a/modules.d/99base/module-setup.sh

+++ b/modules.d/99base/module-setup.sh

@@ -35,7 +35,6 @@ install() {

     inst_script "$moddir/initqueue.sh" "/sbin/initqueue"

     inst_script "$moddir/loginit.sh" "/sbin/loginit"

     inst_script "$moddir/rdsosreport.sh" "/sbin/rdsosreport"

-    inst_script "$moddir/memtrace-ko.sh" "/sbin/tracekomem"

     [ -e "${initdir}/lib" ] || mkdir -m 0755 -p ${initdir}/lib

     mkdir -m 0755 -p ${initdir}/lib/dracut