* [PATCH] perf symbols: Fix symbol size calculation condition
@ 2022-03-17 13:55 Michael Petlan
2022-03-17 15:37 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 3+ messages in thread
From: Michael Petlan @ 2022-03-17 13:55 UTC (permalink / raw)
To: linux-kernel; +Cc: linux-perf-users, acme, jolsa, atrajeev, maddy, kjain
Before this patch, the symbol end address fixup to be called, needed two
conditions being met:
if (prev->end == prev->start && prev->end != curr->start)
Where
"prev->end == prev->start" means that prev is zero-long
(and thus needs a fixup)
and
"prev->end != curr->start" means that fixup hasn't been applied yet
However, this logic is incorrect in the following situation:
*curr = {rb_node = {__rb_parent_color = 278218928,
rb_right = 0x0, rb_left = 0x0},
start = 0xc000000000062354,
end = 0xc000000000062354, namelen = 40, type = 2 '\002',
binding = 0 '\000', idle = 0 '\000', ignore = 0 '\000',
inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
name = 0x1159739e "kprobe_optinsn_page\t[__builtin__kprobes]"}
*prev = {rb_node = {__rb_parent_color = 278219041,
rb_right = 0x109548b0, rb_left = 0x109547c0},
start = 0xc000000000062354,
end = 0xc000000000062354, namelen = 12, type = 2 '\002',
binding = 1 '\001', idle = 0 '\000', ignore = 0 '\000',
inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
name = 0x1095486e "optinsn_slot"}
In this case, prev->start == prev->end == curr->start == curr->end,
thus the condition above thinks that "we need a fixup due to zero
length of prev symbol, but it has been probably done, since the
prev->end == curr->start", which is wrong.
After the patch, the execution path proceeds to arch__symbols__fixup_end
function which fixes up the size of prev symbol by adding page_size to
its end offset.
Signed-off-by: Michael Petlan <mpetlan@redhat.com>
---
tools/perf/util/symbol.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index fb075aa8f8f2..dea0fc495185 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -231,7 +231,7 @@ void symbols__fixup_end(struct rb_root_cached *symbols)
prev = curr;
curr = rb_entry(nd, struct symbol, rb_node);
- if (prev->end == prev->start && prev->end != curr->start)
+ if (prev->end == prev->start || prev->end != curr->start)
arch__symbols__fixup_end(prev, curr);
}
--
2.18.4
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] perf symbols: Fix symbol size calculation condition
2022-03-17 13:55 [PATCH] perf symbols: Fix symbol size calculation condition Michael Petlan
@ 2022-03-17 15:37 ` Arnaldo Carvalho de Melo
2022-03-17 17:49 ` Michael Petlan
0 siblings, 1 reply; 3+ messages in thread
From: Arnaldo Carvalho de Melo @ 2022-03-17 15:37 UTC (permalink / raw)
To: Michael Petlan
Cc: linux-kernel, linux-perf-users, jolsa, atrajeev, maddy, kjain
Em Thu, Mar 17, 2022 at 02:55:36PM +0100, Michael Petlan escreveu:
> Before this patch, the symbol end address fixup to be called, needed two
> conditions being met:
>
> if (prev->end == prev->start && prev->end != curr->start)
>
> Where
> "prev->end == prev->start" means that prev is zero-long
> (and thus needs a fixup)
> and
> "prev->end != curr->start" means that fixup hasn't been applied yet
>
> However, this logic is incorrect in the following situation:
>
> *curr = {rb_node = {__rb_parent_color = 278218928,
> rb_right = 0x0, rb_left = 0x0},
> start = 0xc000000000062354,
> end = 0xc000000000062354, namelen = 40, type = 2 '\002',
> binding = 0 '\000', idle = 0 '\000', ignore = 0 '\000',
> inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
> name = 0x1159739e "kprobe_optinsn_page\t[__builtin__kprobes]"}
>
> *prev = {rb_node = {__rb_parent_color = 278219041,
> rb_right = 0x109548b0, rb_left = 0x109547c0},
> start = 0xc000000000062354,
> end = 0xc000000000062354, namelen = 12, type = 2 '\002',
> binding = 1 '\001', idle = 0 '\000', ignore = 0 '\000',
> inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
> name = 0x1095486e "optinsn_slot"}
>
> In this case, prev->start == prev->end == curr->start == curr->end,
> thus the condition above thinks that "we need a fixup due to zero
> length of prev symbol, but it has been probably done, since the
> prev->end == curr->start", which is wrong.
>
> After the patch, the execution path proceeds to arch__symbols__fixup_end
> function which fixes up the size of prev symbol by adding page_size to
> its end offset.
You missed this:
Fixes: 3b01a413c196c910 ("perf symbols: Improve kallsyms symbol end addr calculation")
I'll add it to the patch, ok?
- Arnaldo
> Signed-off-by: Michael Petlan <mpetlan@redhat.com>
> ---
> tools/perf/util/symbol.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
> index fb075aa8f8f2..dea0fc495185 100644
> --- a/tools/perf/util/symbol.c
> +++ b/tools/perf/util/symbol.c
> @@ -231,7 +231,7 @@ void symbols__fixup_end(struct rb_root_cached *symbols)
> prev = curr;
> curr = rb_entry(nd, struct symbol, rb_node);
>
> - if (prev->end == prev->start && prev->end != curr->start)
> + if (prev->end == prev->start || prev->end != curr->start)
> arch__symbols__fixup_end(prev, curr);
> }
>
> --
> 2.18.4
--
- Arnaldo
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] perf symbols: Fix symbol size calculation condition
2022-03-17 15:37 ` Arnaldo Carvalho de Melo
@ 2022-03-17 17:49 ` Michael Petlan
0 siblings, 0 replies; 3+ messages in thread
From: Michael Petlan @ 2022-03-17 17:49 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: linux-kernel, linux-perf-users, jolsa, atrajeev, maddy, kjain
On Thu, 17 Mar 2022, Arnaldo Carvalho de Melo wrote:
> Em Thu, Mar 17, 2022 at 02:55:36PM +0100, Michael Petlan escreveu:
> > Before this patch, the symbol end address fixup to be called, needed two
> > conditions being met:
> >
> > if (prev->end == prev->start && prev->end != curr->start)
> >
> > Where
> > "prev->end == prev->start" means that prev is zero-long
> > (and thus needs a fixup)
> > and
> > "prev->end != curr->start" means that fixup hasn't been applied yet
> >
> > However, this logic is incorrect in the following situation:
> >
> > *curr = {rb_node = {__rb_parent_color = 278218928,
> > rb_right = 0x0, rb_left = 0x0},
> > start = 0xc000000000062354,
> > end = 0xc000000000062354, namelen = 40, type = 2 '\002',
> > binding = 0 '\000', idle = 0 '\000', ignore = 0 '\000',
> > inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
> > name = 0x1159739e "kprobe_optinsn_page\t[__builtin__kprobes]"}
> >
> > *prev = {rb_node = {__rb_parent_color = 278219041,
> > rb_right = 0x109548b0, rb_left = 0x109547c0},
> > start = 0xc000000000062354,
> > end = 0xc000000000062354, namelen = 12, type = 2 '\002',
> > binding = 1 '\001', idle = 0 '\000', ignore = 0 '\000',
> > inlined = 0 '\000', arch_sym = 0 '\000', annotate2 = false,
> > name = 0x1095486e "optinsn_slot"}
> >
> > In this case, prev->start == prev->end == curr->start == curr->end,
> > thus the condition above thinks that "we need a fixup due to zero
> > length of prev symbol, but it has been probably done, since the
> > prev->end == curr->start", which is wrong.
> >
> > After the patch, the execution path proceeds to arch__symbols__fixup_end
> > function which fixes up the size of prev symbol by adding page_size to
> > its end offset.
>
> You missed this:
>
> Fixes: 3b01a413c196c910 ("perf symbols: Improve kallsyms symbol end addr calculation")
>
> I'll add it to the patch, ok?
Sure thing. Thanks.
M
>
> - Arnaldo
>
> > Signed-off-by: Michael Petlan <mpetlan@redhat.com>
> > ---
> > tools/perf/util/symbol.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
> > index fb075aa8f8f2..dea0fc495185 100644
> > --- a/tools/perf/util/symbol.c
> > +++ b/tools/perf/util/symbol.c
> > @@ -231,7 +231,7 @@ void symbols__fixup_end(struct rb_root_cached *symbols)
> > prev = curr;
> > curr = rb_entry(nd, struct symbol, rb_node);
> >
> > - if (prev->end == prev->start && prev->end != curr->start)
> > + if (prev->end == prev->start || prev->end != curr->start)
> > arch__symbols__fixup_end(prev, curr);
> > }
> >
> > --
> > 2.18.4
>
> --
>
> - Arnaldo
>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-03-17 17:49 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-03-17 13:55 [PATCH] perf symbols: Fix symbol size calculation condition Michael Petlan
2022-03-17 15:37 ` Arnaldo Carvalho de Melo
2022-03-17 17:49 ` Michael Petlan
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).