In https://lore.kernel.org/lkml/20211209150910.GA23668@axis.com/
Vincent's patch commented on, and worked around, a bug toggling
static_branch's, when a 2nd PRINTK-ish flag was added. The bug
results in a premature static_branch_disable when the 1st of 2 flags
was disabled.
The cited commit computed newflags, but then in the JUMP_LABEL block,
failed to use that result, instead using just one of the terms in it.
Using newflags instead made the code work properly.
This is Vincents test-case, reduced. It needs the 2nd flag to
demonstrate the bug, but it's explanatory here.
pt_test() {
echo 5 > /sys/module/dynamic_debug/verbose
site="module tcp" # just one callsite
echo " $site =_ " > /proc/dynamic_debug/control # clear it
# A B ~A ~B
for flg in +T +p "-T #broke here" -p; do
echo " $site $flg " > /proc/dynamic_debug/control
done;
# A B ~B ~A
for flg in +T +p "-p #broke here" -T; do
echo " $site $flg " > /proc/dynamic_debug/control
done
}
pt_test
Fixes: 84da83a6ff dyndbg: combine flags & mask into a struct, simplify with it
CC: vincent.whitchurch@axis.com
Acked-by: Jason Baron <jbaron@akamai.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20220904214134.408619-2-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
adjust current v*pr_info() calls to fit an overview..detail scheme:
1- module level activity: add/remove, etc
2- command ingest, splitting, summary of effects.
per >control write
3- command parsing: op, flags, search terms
4- per-site change msg
can yield ~3k x 2 logs per echo "+p;-p" > command.
Summarize these 4 levels in MODULE_PARM_DESC, and update verbose=3 in Doc.
2- is new, to isolate a problem where a stress-test script (which
feeds ~4kb multi-command strings) would produce short writes,
truncating last command and causing parsing errors, which confused
test results. The script fix was to use syswrite, to deliver full
proper commands.
4- gets per-callsite "changed:" pr-infos, which are very noisy during
stress tests, and formerly obscured v1-3 messages, and overwhelmed the
static-key workload being tested.
The verbose parameter has previously seen adjustment:
commit 481c0e33f1 ("dyndbg: refine debug verbosity; 1 is basic, 2 more chatty")
The script driving these adjustments is:
!/usr/bin/perl -w
=for Doc
1st purpose was to benchmark the effect of wildcard queries on query
performance; if wildcards are risk free cheap enough, we can deploy
them in the (floating) format search. 1st finding: wildcards take 2x
as long to process.
2nd purpose was to benchmark real static-key changes VS simple flag
changes. Found ~100x decrease for the hard work.
The script maximizes workload per >control by packing it a ~4kb
string of "+p; -p;" commands; this uncovered some broken stuff.
The 85th query failed, and appears to be truncated, so is gramatically
incorrect. Its either an error here, or in the kernel. Its not
happening atm, retest.
Plot thickens: fail only happens doing +-p, not +-mf, likely load
dependent. Error remains consistent. Looks like a short write,
longer on writer than kernel-reader. Try syswrite on handle to
control this. That fixed short write.
=cut
use Getopt::Std;
getopts('vN:k:', \my %opts) or die <<EOH;
$0 options:
-v verbose
-k=n kernel dyndbg verbosity
-N=n number of loops.. tbrc
EOH
$opts{N} //= 10; # !undef, 0 tests too long.
my $ctrl = '/proc/dynamic_debug/control';
vx($opts{k}) if defined $opts{k}; # works on -k0
open(my $CTL, '>', $ctrl) or die "cant open $ctrl for writing: $!\n";
sub vx {
my $arg = shift;
my $cmd = "echo $arg > /sys/module/dynamic_debug/parameters/verbose";
system($cmd);
warn("vx problem: rc:$? err:$! qry: $cmd\n") if ($?);
}
sub qryOK {
my $qry = shift;
print "syntax test: <\n$qry>\n" if $opts{v};
my $bytes = syswrite $CTL, $qry;
printf "short read: $bytes / %d\n", length $qry if $bytes < length $qry;
if ($?) {
warn "rc:$? err:$! qry: $qry\n";
return 0;
}
return 1;
}
sub build_queries {
my ($cmd, $flags, $ct) = @_;
# build experiment and reference queries
my $cycle = " $cmd +$flags # on ; $cmd -$flags # off \n";
my $ref = " +$flags ; -$flags \n";
my $len = length $cycle;
my $max = int(4096 / $len); # break/fit to buffer size
$ct |= $max;
print "qry: ct:$max x << \n$cycle >>\n";
return unless qryOK($ref);
return unless qryOK($cycle);
my $wild = $cycle x $ct;
my $empty = $ref x $ct;
printf "len: %d, %d\n", length $wild, length $empty;
return { trial => $wild,
ref => $empty,
probe => $cycle,
zero => $ref,
count => $ct,
max => $max
};
}
my $query_set = build_queries(' file "*" module "*" func "*" ', "mf");
qryOK($query_set->{zero});
qryOK($query_set->{probe});
qryOK($query_set->{ref});
qryOK($query_set->{trial});
use Benchmark;
sub dobatch {
my ($cmd, $flags, $reps, $ct) = @_;
$reps ||= $opts{N};
my $qs = build_queries($cmd, $flags, $ct);
timethese($reps,
{
wildcards => sub {
syswrite $CTL, $qs->{trial};
},
no_search => sub {
syswrite $CTL, $qs->{ref};
}
}
);
}
sub bench_static_key_toggle {
vx 0;
dobatch(' file "*" module "*" func "*" ', "mf");
dobatch(' file "*" module "*" func "*" ', "p");
}
sub bench_verbose_levels {
for my $i (0..4) {
vx $i;
dobatch(' file "*" module "*" func "*" ', "mf");
}
}
bench_static_key_toggle();
__END__
Heres how the test-script runs:
:: verbose=3 parsing info
[ 48.401646] dyndbg: query 95: "file "*" module "*" func "*" -mf # off " mod:*
[ 48.402040] dyndbg: split into words: "file" "*" "module" "*" "func" "*" "-mf"
[ 48.402456] dyndbg: op='-'
[ 48.402615] dyndbg: flags=0x6
[ 48.402779] dyndbg: *flagsp=0x0 *maskp=0xfffffff9
[ 48.403033] dyndbg: parsed: func="*" file="*" module="*" format="" lineno=0-0
[ 48.403674] dyndbg: applied: func="*" file="*" module="*" format="" lineno=0-0
:: verbose=2 >control summary.
~300k site matches/changes per 4kb command
[ 48.404063] dyndbg: processed 96 queries, with 296160 matches, 0 errs
:: 2 queries against each other, no-search vs all-wildcard-search
qry: ct:48 x <<
file "*" module "*" func "*" +mf # on ; file "*" module "*" func "*" -mf # off
>>
len: 4080, 576
Benchmark: timing 10 iterations of no_search, wildcards...
no_search: 0 wallclock secs ( 0.00 usr + 0.03 sys = 0.03 CPU) @ 333.33/s (n=10)
(warning: too few iterations for a reliable count)
wildcards: 0 wallclock secs ( 0.00 usr + 0.09 sys = 0.09 CPU) @ 111.11/s (n=10)
(warning: too few iterations for a reliable count)
:: 2 queries, both doing real work / changing stati-key states.
qry: ct:49 x <<
file "*" module "*" func "*" +p # on ; file "*" module "*" func "*" -p # off
>>
len: 4067, 490
Benchmark: timing 10 iterations of no_search, wildcards...
no_search: 20 wallclock secs ( 0.00 usr + 20.36 sys = 20.36 CPU) @ 0.49/s (n=10)
wildcards: 21 wallclock secs ( 0.00 usr + 21.08 sys = 21.08 CPU) @ 0.47/s (n=10)
bash-5.1#
Thats 150k static-key-toggles / sec
~600x slower than simple flags
on qemu --smp 3 run
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20211019210746.185307-1-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The cited commit inadvertently altered the verbose level of a
vpr_info, restore it to original.
Fixes: 216a0fc408 ("dyndbg: show module in vpr-info in dd-exec-queries")
Signed-off-By: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20211014223614.1952171-1-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
when `echo $cmd > control` contains multiple queries, extra query
separators (;\n) can parse as empty statements. This is normal, and a
vpr-info on an empty command is just noise.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20211013220726.1280565-4-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On qemu --smp 3 runs, remove-module can get called 3 times.
So don't print on entry; instead print "removed" after entry is
found and removed, so just once.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20211013220726.1280565-3-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This param has been deprecated for a very long time now, let's rip it
out.
Signed-off-by: Andrew Halaney <ahalaney@redhat.com>
Signed-off-by: Jason Baron <jbaron@akamai.com>
Link: https://lore.kernel.org/r/1634139622-20667-3-git-send-email-jbaron@akamai.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Right now dyndbg shows up as an unknown parameter if used on boot:
Unknown command line parameters: dyndbg=+p
That's because it is unknown, it doesn't sit in the __param
section, so the processing done to warn users supplying an unknown
parameter doesn't think it is legitimate.
Install a dummy handler to register it. dynamic debug needs to search
the whole command line for modules listed that are currently builtin,
so there's no real work to be done in this callback.
Fixes: 86d1919a4f ("init: print out unknown kernel parameters")
Tested-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Andrew Halaney <ahalaney@redhat.com>
Signed-off-by: Jason Baron <jbaron@akamai.com>
Link: https://lore.kernel.org/r/1634139622-20667-2-git-send-email-jbaron@akamai.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
dynamic_debug_exec_queries() accepts a separate module arg (so it can
support $module.dyndbg boot arg), display that in the vpr-info for a
more useful user-debug context.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20211012183310.1016678-2-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If booted with verbose>=1, dyndbg prints the memory usage in bytes,
of builtin modules' prdebugs. KiB reads better.
no functional changes.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20210525033240.35260-1-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove a vpr_info which I added in 2012, when I knew even less than now.
In 2020, a simpler pr_fmt stripped it of context, and any remaining value.
no functional change.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20210504222235.1033685-3-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Wrap function in a static-inline one, which checks flags to avoid
calling the function unnecessarily.
And hoist its output-buffer initialization to the grand-caller, which
is already allocating the buffer on the stack, and can trivially
initialize it too.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20210504222235.1033685-2-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Query like 'file tcp_input.c line 1234 +p' was broken by
commit aaebe329bf ("dyndbg: accept 'file foo.c:func1' and 'file
foo.c:10-100'") because a file name without a ':' now makes the loop in
ddebug_parse_query() exits early before parsing the 'line 1234' part.
As a result, all pr_debug() in tcp_input.c will be enabled, instead of only
the one on line 1234. Changing 'break' to 'continue' fixes this.
Fixes: aaebe329bf ("dyndbg: accept 'file foo.c:func1' and 'file foo.c:10-100'")
Cc: stable <stable@vger.kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Shuo Chen <shuochen@google.com>
Acked-by: Jason Baron <jbaron@akamai.com>
Link: https://lore.kernel.org/r/20210414212400.2927281-1-giantchen@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In commit a2d375eda7 ("dyndbg: refine export, rename to
dynamic_debug_exec_queries()"), a string is copied before checking it
isn't NULL. Fix this, report a usage/interface error, and return the
proper error code.
Fixes: a2d375eda7 ("dyndbg: refine export, rename to dynamic_debug_exec_queries()")
Cc: stable@vger.kernel.org
--
-v2 drop comment tweak, improve commit message
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20201209183625.2432329-1-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit 14775b0496 as there
were still some parsing problems with it, and the follow-on patch for
it.
Let's revisit it later, just drop it for now.
Cc: <jbaron@akamai.com>
Cc: Jim Cromie <jim.cromie@gmail.com>
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Fixes: 14775b0496 ("dyndbg: accept query terms like file=bar and module=foo")
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit 42f07816ac as it
still causes problems. It will be resolved later, let's revert it so we
can also revert the original patch this was supposed to be helping with.
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Fixes: 42f07816ac ("dyndbg: fix problem parsing format="foo bar"")
Cc: Jim Cromie <jim.cromie@gmail.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 14775b0496 ("dyndbg: accept query terms like file=bar and
module=foo") added the combined keyword=value parsing poorly; revert
most of it, keeping the keyword & arg change.
Instead, fix the tokenizer for the new input, by terminating the
keyword (an unquoted word) on '=' as well as space, thus letting the
tokenizer work on the quoted argument, like it would have previously.
Also add a few debug-prints to show more parsing context, into
tokenizer and parse-query, and use "keyword, value" in others.
Fixes: 14775b0496 ("dyndbg: accept query terms like file=bar and module=foo")
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200831182210.850852-4-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4c0d77828d ("dyndbg: export ddebug_exec_queries") had a few
problems:
- broken non DYNAMIC_DEBUG_CORE configs, sparse warning
- the exported function modifies query string, breaks on RO strings.
- func name follows internal convention, shouldn't be exposed as is.
1st is fixed in header with ifdefd function prototype or stub defn.
Also remove an obsolete HAVE-symbol ifdef-comment, and add others.
Fix others by wrapping existing internal function with a new one,
named in accordance with module-prefix naming convention, before
export hits v5.9.0. In new function, copy query string to a local
buffer, so users can pass hard-coded/RO queries, and internal function
can be used unchanged.
Fixes: 4c0d77828d ("dyndbg: export ddebug_exec_queries")
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200831182210.850852-3-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Export ddebug_exec_queries() for use by modules.
This will allow module authors to control all their *pr_debug*s
dynamically. And since ddebug_exec_queries() is what implements
"echo $query >control", it gives the same per-callsite control.
Virtues of this:
- simplicity. just an export.
- full control over any/all subsets of callsites.
- same "query/command-string" in code and console
- full callsite selectivity with module file line format
Format in particular deserves special attention; it is where
low-hanging fruit will be found.
Consider: drivers/gpu/drm/amd/display/include/logger_types.h:
#define DC_LOG_SURFACE(...) pr_debug("[SURFACE]:"__VA_ARGS__)
#define DC_LOG_HW_LINK_TRAINING(...) pr_debug("[HW_LINK_TRAINING]:"__VA_ARGS__)
.. 9 more ..
Thats 11 string prefixes, used in 804 places in drivers/gpu/**
Clearly this is a systematized classification of those callsites.
And one I'd expect to see repeated often.
Using ddebug_exec_queries(), authors can select on those prefixes
as a unitary set, equivalent to:
echo "module=MODULE_NAME format=^[SURFACE]: +p" >control
Trivially, those sets can be subsected with the other query terms too,
say file=foo, should the author see fit.
Perhaps as important, users can modify the set of enabled callsites,
presumably to aid debugging by enabling helpful debug callsites, and
disabling those that just clutter the info.
Authors could even alter [fmlt] flags, though I dont see a good reason
why they would. Perhaps harnessed by bug-logging automation to get
fuller, or more minimal bug-reports.
DRM
drm has both drm.debug, which defines 32 categories of drm_printk
logging, and entirely separate uses of pr_debug, which are dynamic on
this i915 laptop, running mainline. So I can observe and report on
both.
The i915 driver has 118 dyndbg callsites, with following
"classifications" defined in drivers/gpu/drm/i915/gvt/**
$ grep 915 /proc/dynamic_debug/control | cut -d= -f2 | cut -d: -f1,2 | sort -u
_ "gvt: cmd
_ "gvt: core
_ "gvt: dpy
_ "gvt: el
_ "gvt: irq
_ "gvt: mm
_ "gvt: mmio
_ "gvt: render
_ "gvt: sched
_ "%s for root hub!\012"
_ "Vendor defined info completion code %u\012"
This classification is entirely out-of-band for control by drm.debug,
and is only available to root user at the console. But module authors
can activate them with ddebug_exec_queries(sprintf("format=^%s +p")),
and then decide how to expose the groups to the user for max utility.
drm.debug
drm.debug has 32 bit-flags, and matching enum drm_debug_category
values to classify the ~2943 DRM_DEBUG*() callsites in drivers/gpu
The drm.debug callback could invoke ddebug_exec_queries() with 32
different hardcoded query strings, needing only (bit) ? " +p" : " -p"
added.
I briefly enabled drm.debug=0xff on my i915 laptop, which yielded
these unique prefixes: (dmesg | cut -c17- | cut -d\] -f1 | sort -u)
[drm:drm_atomic_check_only [drm
[drm:drm_atomic_get_crtc_state [drm
[drm:drm_atomic_get_plane_state [drm
[drm:drm_atomic_nonblocking_commit [drm
[drm:drm_atomic_set_fb_for_plane [drm
[drm:drm_atomic_state_default_clear [drm
[drm:__drm_atomic_state_free [drm
[drm:drm_atomic_state_init [drm
[drm:drm_crtc_vblank_helper_get_vblank_timestamp_internal [drm
[drm:drm_handle_vblank [drm
[drm:drm_ioctl [drm
[drm:drm_mode_addfb2 [drm
[drm:drm_mode_object_get [drm
[drm:drm_mode_object_put.part.0 [drm
[drm:drm_update_vblank_count [drm
[drm:drm_vblank_enable [drm
[drm:drm_vblank_restore [drm
[drm:vblank_disable_fn [drm
i915 0000:00:02.0: [drm:gen9_set_dc_state [i915
i915 0000:00:02.0: [drm:intel_atomic_get_global_obj_state [i915
i915 0000:00:02.0: [drm:__intel_display_power_get_domain.part.0 [i915
i915 0000:00:02.0: [drm:__intel_display_power_put_domain [i915
i915 0000:00:02.0: [drm:intel_plane_atomic_calc_changes [i915
i915 0000:00:02.0: [drm:skl_enable_dc6 [i915
Several good format=^prefixes are apparent there, and some misses.
^[drm:drm_atomic_ # misses: [drm:__drm_atomic_state_free [drm
^[drm:drm_ioctl
^[drm:drm_mode
^[drm:drm_vblank_ # misses: [drm:drm_update_vblank_count & [drm:vblank_disable_fn
Its not a perfect 1:1 single format-match per class, but the misses
above can be covered with 1 & 2 additional queries, which can be
concatenated together with ";" separators and submitted with 1 call.
Benefits:
For drm, adapting DRM_DEBUG to use dynamic-debug inside could
replicate (and thereby obsolete) lots of bit-checking in current
DRM_DEBUG callsites, at least with JUMP_LABEL optimized code.
ddebug_exec_queries() and a handful of fixed query-strings can select
and thereby control the already classified callsites.
With the classes mapped to queries, the enum type and parameter can be
eliminated (folded away with macro magic), at least for DYNAMIC_DEBUG
& JUMP_LABEL builds.
Is it safe ?
ddebug_exec_queries() is currently exposed to user space in
several limited ways;
1 it is called from module-load callback, where it implements the
$modname.dyndbg=+p "fake" parameter provided to all modules.
2 it handles query input via >control directly
IOW, it is "fully" exposed to local root user; exposing the same
functionality to other kernel modules is no additional risk.
The other standard issue to check is locking:
dyndbg has a single mutex, taken by ddebug_change to handle >control,
and by ddebug_proc_(start|stop) to span `cat control`. Queries
submitted via export will typically have module specified, which
dramatically cuts the scan by ddebug_change vs "module=* +p".
ISTM this proposed export presents no locking problems.
TLDR;
It would be interesting to see how drm.dyndbg=$QUERY and
drm.debug=$HEXY would interact; it might be order dependent, as
if given as modprobe args or in /etc/modprobe.d/
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-19-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For log-message output, reduce column space consumed by current
pr_fmt by dropping __func__ and shortening "dynamic_debug" to
"dyndbg". This improves readability on narrow consoles, and better
matches other kernel boot info messages.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-18-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This should work:
echo module=amd* format=^[IF_TRACE]: +p >/proc/dynamic_debug/control
consider drivers/gpu/drm/amd/display/include/logger_types.h:
It has 11 defines like:
#define DC_LOG_IF_TRACE(...) pr_debug("[IF_TRACE]:"__VA_ARGS__)
These defines are used 804 times at recent count; they are a good use
case to evaluate existing format-message based classifications of
*pr_debug*. Those macros prefix the supplied format with a fixed
string, I'd expect most existing message classification schemes to do
something similar.
Hence we want to be able to anchor our match to the beginning of the
format string, allowing easy construction of clear and precise
queries, leveraging the existing classification scheme to enable and
disable those callsites.
Note that unlike other search terms, formats are implicitly floating
substring matches, without the need for explicit wildcards.
This makes no attempt at wider regex features, just the one we need.
TLDR: Using the anchor also means the []s are less helpful for
disamiguating the prefix from a random in-message occurrence, allowing
shorter prefixes.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-17-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
flags & mask are used together everywhere, and are passed around
together between multiple functions; they belong together in a struct,
call that struct flag_settings.
Use struct flag_settings to rework 3 functions:
- ddebug_exec_query - declares query and flag-settings,
calls other 2, passing flags
- ddebug_parse_flags - fills flag_settings and returns
- ddebug_change - test all callsites against query,
modify passing sites.
benefits:
- bit-banging always needs flags & mask, best together.
- simpler function signatures
- 1 less parameter, less stack overhead
no functional changes
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-16-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Current code expects "keyword" "arg" as 2 words, space separated.
Change to also accept "keyword=arg" form as well, and drop !(nwords%2)
requirement. Then in rest of function, use new keyword, arg variables
instead of word[i], word[i+1]
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-15-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Accept these additional query forms:
echo "file $filestr +_" > control
path/to/file.c:100 # as from control, column 1
path/to/file.c:1-100 # or any legal line-range
path/to/file.c:func_A # as from an editor/browser
path/to/file.c:drm_* # wildcards still work
path/to/file.c:*_foo # lead wildcard too
1st 2 examples are treated as line-ranges, 3-5 are treated as func's
Doc these changes, and sprinkle in a few extra wild-card examples and
trailing # explanation texts.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-14-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Make the code-block reusable to later handle "file foo.c:101-200" etc.
This is a 99% code move, with reindent, function wrap&call, +pr_debug.
no functional changes.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-13-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
loadable modules are the last in on this list, and are the only
modules that could be removed. ddebug_remove_module() searches from
head, but ddebug_add_module() uses list_add_tail(). Change it to
list_add() for a micro-optimization.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-11-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ddebug_exec_query declares an auto var, and passes it to
ddebug_parse_query, which memsets it before using it. Drop that
memset, instead initialize the variable in the caller; let the
compiler decide how to do it.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-10-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
this pr_err attempts to print the string after the OP, but the string
has been parsed and chopped up, so looks empty.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-9-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ddebug_describe_flags() currently fills a caller provided string buffer,
after testing its size (also passed) in a BUG_ON. Fix this by
replacing them with a known-big-enough string buffer wrapped in a
struct, and passing that instead.
Also simplify ddebug_describe_flags() flags parameter from a struct to
a member in that struct, and hoist the member deref up to the caller.
This makes the function reusable (soon) where flags are unpacked.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-8-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
during dyndbg init, verbose logging prints its ram overhead. It
counted strlens of struct _ddebug's 4 string members, in all callsite
entries, which would be approximately correct if each had been
mallocd. But they are pointers into shared .rodata; for example, all
10 kobject callsites have identical filename, module values.
Its best not to count that memory at all, since we cannot know they
were linked in because of CONFIG_DYNAMIC_DEBUG=y, and we want to
report a number that reflects what ram is saved by deconfiguring it.
Also fix wording and size under-reporting of the __dyndbg section.
Heres my overhead, on a virtme-run VM on a fedora-31 laptop:
dynamic_debug:dynamic_debug_init: 260 modules, 2479 entries \
and 10400 bytes in ddebug tables, 138824 bytes in __dyndbg section
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-7-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
dyndbg populates its callsite info into __verbose section, change that
to a more specific and descriptive name, __dyndbg.
Also, per checkpatch:
simplify __attribute(..) to __section(__dyndbg) declaration.
and 1 spelling fix, decriptor
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-6-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The verbose/debug logging done for `cat $MNT/dynamic_debug/control` is
voluminous (2 per control file entry + 2 per PAGE). Moreover, it just
prints pointer and sequence, which is not useful to a dyndbg user.
So just drop them.
Also require verbose>=2 for several other debug printks that are a bit
too chatty for typical needs;
ddebug_change() prints changes, once per modified callsite. Since
queries like "+p" will enable ~2300 callsites in a typical laptop, a
user probably doesn't need to see them often. ddebug_exec_queries()
still summarizes with verbose=1.
ddebug_(add|remove)_module() also print 1 line per action on a module,
not needed by typical modprobe user.
This leaves verbose=1 better focussed on the >control parsing process.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-5-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4bad78c550 ("lib/dynamic_debug.c: use seq_open_private() instead of seq_open()")'
The commit was one of a tree-wide set which replaced open-coded
boilerplate with a single tail-call. It therefore obsoleted the
comment about that boilerplate, clean that up now.
Acked-by: <jbaron@akamai.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20200719231058.1586423-4-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Instead of enabling dynamic debug globally with CONFIG_DYNAMIC_DEBUG,
CONFIG_DYNAMIC_DEBUG_CORE will only enable core function of dynamic
debug. With the DYNAMIC_DEBUG_MODULE defined for any modules, dynamic
debug will be tied to them.
This is useful for people who only want to enable dynamic debug for
kernel modules without worrying about kernel image size and memory
consumption is increasing too much.
[orson.zhai@unisoc.com: v2]
Link: http://lkml.kernel.org/r/1587408228-10861-1-git-send-email-orson.unisoc@gmail.com
Signed-off-by: Orson Zhai <orson.zhai@unisoc.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Petr Mladek <pmladek@suse.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Jason Baron <jbaron@akamai.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Link: http://lkml.kernel.org/r/1586521984-5890-1-git-send-email-orson.unisoc@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Clang warns:
../lib/dynamic_debug.c:1034:24: warning: array comparison always
evaluates to false [-Wtautological-compare]
if (__start___verbose == __stop___verbose) {
^
1 warning generated.
These are not true arrays, they are linker defined symbols, which are just
addresses. Using the address of operator silences the warning and does
not change the resulting assembly with either clang/ld.lld or gcc/ld
(tested with diff + objdump -Dr).
Suggested-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Jason Baron <jbaron@akamai.com>
Link: https://github.com/ClangBuiltLinux/linux/issues/894
Link: http://lkml.kernel.org/r/20200220051320.10739-1-natechancellor@gmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
With the realization that having debugfs enabled on "production" systems
is generally not a good idea, debugfs is being disabled from more and
more platforms over time. However, the functionality of dynamic
debugging still is needed at times, and since it relies on debugfs for
its user api, having debugfs disabled also forces dynamic debug to be
disabled.
To get around this, also create the "control" file for dynamic_debug in
procfs. This allows people turn on debugging as needed at runtime for
individual driverfs and subsystems.
Reported-by: many different companies
Cc: Jason Baron <jbaron@akamai.com>
Acked-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20200210211142.GB1373304@kroah.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.
Cc: linux-kernel@vger.kernel.org
Acked-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This serves two purposes: First, we get a diagnostic if (though
extremely unlikely), any of the calls of ddebug_add_module for built-in
code fails, effectively disabling dynamic_debug. Second, I want to make
struct _ddebug opaque, and avoid accessing any of its members outside
dynamic_debug.[ch].
Link: http://lkml.kernel.org/r/20190212214150.4807-9-linux@rasmusvillemoes.dk
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Acked-by: Jason Baron <jbaron@akamai.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: "Rafael J . Wysocki" <rafael.j.wysocki@intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The only caller of ddebug_{add,remove}_module outside dynamic_debug.c is
kernel/module.c, which is obviously not itself modular (though it would
be an interesting exercise to make that happen...). I also fail to see
how these interfaces can be used by modules, in-tree or not.
Link: http://lkml.kernel.org/r/20190212214150.4807-8-linux@rasmusvillemoes.dk
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Acked-by: Jason Baron <jbaron@akamai.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: "Rafael J . Wysocki" <rafael.j.wysocki@intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that we store the passed-in string directly in ddebug_add_module, we
can use pointer equality instead of strcmp. This is a little more
efficient, but more importantly, this also makes the code somewhat more
correct:
Currently, if one loads and then unloads a module whose name happens to
match the KBUILD_MODNAME of some built-in functionality (which need not
even be modular at all), all of their dynamic debug entries vanish along
with those of the actual module. For example, loading and unloading a
core.ko hides all pr_debugs from drivers/base/core.c and other built-in
files called core.c (incidentally, there is an in-tree module whose name
is core, but I just tested this with an out-of-tree trivial one).
Link: http://lkml.kernel.org/r/20190212214150.4807-7-linux@rasmusvillemoes.dk
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Acked-by: Jason Baron <jbaron@akamai.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: "Rafael J . Wysocki" <rafael.j.wysocki@intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For built-in modules, we're already reusing the passed-in string via
kstrdup_const(). But for actual modules (i.e. when we're called from
dynamic_debug_setup in module.c), the passed-in string (which points at
the name[] array inside struct module) is also guaranteed to live at
least as long as the struct ddebug_table, since free_module() calls
ddebug_remove_module().
Link: http://lkml.kernel.org/r/20190212214150.4807-6-linux@rasmusvillemoes.dk
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Acked-by: Jason Baron <jbaron@akamai.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: "Rafael J . Wysocki" <rafael.j.wysocki@intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, CONFIG_JUMP_LABEL just means "I _want_ to use jump label".
The jump label is controlled by HAVE_JUMP_LABEL, which is defined
like this:
#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
# define HAVE_JUMP_LABEL
#endif
We can improve this by testing 'asm goto' support in Kconfig, then
make JUMP_LABEL depend on CC_HAS_ASM_GOTO.
Ugly #ifdef HAVE_JUMP_LABEL will go away, and CONFIG_JUMP_LABEL will
match to the real kernel capability.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
line-range is supposed to treat "1-" as "1-endoffile", so
handle the special case by setting last_lineno to UINT_MAX.
Fixes this error:
dynamic_debug:ddebug_parse_query: last-line:0 < 1st-line:1
dynamic_debug:ddebug_exec_query: query parse failed
Link: http://lkml.kernel.org/r/10a6a101-e2be-209f-1f41-54637824788e@infradead.org
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Although dynamic debug is often only used for debug builds, sometimes
its enabled for production builds as well. Minimize its impact by using
jump labels. This reduces the text section by 7000+ bytes in the kernel
image below. It does increase data, but this should only be referenced
when changing the direction of the branches, and hence usually not in
cache.
text data bss dec hex filename
8194852 4879776 925696 14000324 d5a0c4 vmlinux.pre
8187337 4960224 925696 14073257 d6bda9 vmlinux.post
Link: http://lkml.kernel.org/r/d165b465e8c89bc582d973758d40be44c33f018b.1467837322.git.jbaron@akamai.com
Signed-off-by: Jason Baron <jbaron@akamai.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Joe Perches <joe@perches.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A _lot_ of ->write() instances were open-coding it; some are
converted to memdup_user_nul(), a lot more remain...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Using kstrdup_const, thus reusing .rodata when possible, saves around 2 kB
of runtime memory on my laptop/.config combination.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Jason Baron <jbaron@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
