* [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
@ 2024-08-26 12:37 Sergey Kaplun via Tarantool-patches
2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2024-08-26 12:37 UTC (permalink / raw)
To: Maxim Kokryashkin, Sergey Bronnikov; +Cc: tarantool-patches
From: Mike Pall <mike>
Thanks to Sergey Kaplun.
(cherry picked from commit e45fd4cb713b610506213692f3b55a1869febb03)
`narrow_conv_backprop()` misses the stack pointer (`nc->sp`) limit check
after a bunch of recursive calls that may change its value. As a result,
it leads to stack-buffer-overflow during the instruction narrowing. This
patch adds a missing check.
Sergey Kaplun:
* added the description and the test for the problem
Part of tarantool/tarantool#10199
---
Branch: https://github.com/tarantool/luajit/tree/skaplun/lj-1262-fix-limit-narrow-conv-backprop
Related issues:
* https://github.com/tarantool/tarantool/issues/10199
* https://github.com/LuaJIT/LuaJIT/issues/1262
src/lj_opt_narrow.c | 3 +-
...62-fix-limit-narrow-conv-backprop.test.lua | 61 +++++++++++++++++++
2 files changed, 63 insertions(+), 1 deletion(-)
create mode 100644 test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
diff --git a/src/lj_opt_narrow.c b/src/lj_opt_narrow.c
index db0da10f..6b6f20d3 100644
--- a/src/lj_opt_narrow.c
+++ b/src/lj_opt_narrow.c
@@ -341,7 +341,8 @@ static int narrow_conv_backprop(NarrowConv *nc, IRRef ref, int depth)
NarrowIns *savesp = nc->sp;
int count = narrow_conv_backprop(nc, ir->op1, depth);
count += narrow_conv_backprop(nc, ir->op2, depth);
- if (count <= 1) { /* Limit total number of conversions. */
+ /* Limit total number of conversions. */
+ if (count <= 1 && nc->sp < nc->maxsp) {
*nc->sp++ = NARROWINS(IRT(ir->o, nc->t), ref);
return count;
}
diff --git a/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
new file mode 100644
index 00000000..6bb4025d
--- /dev/null
+++ b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
@@ -0,0 +1,61 @@
+local tap = require('tap')
+
+-- Test file to demonstrate stack-buffer-overflow during the
+-- narrowing optimization.
+-- See also: https://github.com/LuaJIT/LuaJIT/issues/1262.
+
+local test = tap.test('lj-1262-fix-limit-narrow-conv-backprop'):skipcond({
+ ['Test requires JIT enabled'] = not jit.status(),
+})
+
+test:plan(1)
+
+-- XXX: Test fails only under ASAN.
+-- XXX: The original reproducer was found by fuzzer:
+-- https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=70779.
+-- It creates a long side trace with a huge amount of ADD IRs,
+-- which are recursively used in the `narrow_conv_backprop()` many
+-- enough times to catch the stack-buffer-overflow. I can't
+-- simplify the reproducer any more (or write it from scratch), so
+-- I leave it in that state.
+
+local DEFAULT_NUMBER = 1
+local tonumber = tonumber
+
+local always_number = function(val)
+ return tonumber(val) or DEFAULT_NUMBER
+end
+
+local add = function(v1, v2)
+ return always_number(v1) + always_number(v2)
+end
+
+jit.opt.start('hotloop=1', 'hotexit=1')
+
+local counter_0 = 0
+local counter_1 = 0
+local counter_2 = 0
+local tmp = add(nil, 'Name')
+local Name0 = add(tmp, 'Name')
+-- Start a long side trace here.
+for _ = 0, 0, 0 do
+ if counter_0 > 5 then break end
+ counter_0 = counter_0 + 1
+
+ for _ = always_number(false), 1, always_number(Name0) do
+ if counter_1 > 5 then break end
+ counter_1 = counter_1 + 1
+
+ repeat
+ if counter_2 > 5 then break end
+ counter_2 = counter_2 + 1
+
+ Name0 = Name0 + Name0 + Name0
+ Name0 = add(Name0, nil) + Name0
+ until nil
+ end
+end
+
+test:ok(true, 'no stack-buffer-overflow during narrowing')
+
+test:done(true)
--
2.46.0
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
2024-08-26 12:37 [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop() Sergey Kaplun via Tarantool-patches
@ 2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
2024-09-17 12:01 ` Sergey Kaplun via Tarantool-patches
2024-09-23 6:41 ` Maxim Kokryashkin via Tarantool-patches
2024-10-18 15:15 ` Sergey Kaplun via Tarantool-patches
2 siblings, 1 reply; 6+ messages in thread
From: Sergey Bronnikov via Tarantool-patches @ 2024-09-17 10:32 UTC (permalink / raw)
To: Sergey Kaplun, Maxim Kokryashkin; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 4144 bytes --]
Hi, Sergey,
thanks for tha patch! LGTM with a minor comment below.
On 26.08.2024 15:37, Sergey Kaplun wrote:
> From: Mike Pall <mike>
>
> Thanks to Sergey Kaplun.
>
> (cherry picked from commit e45fd4cb713b610506213692f3b55a1869febb03)
>
> `narrow_conv_backprop()` misses the stack pointer (`nc->sp`) limit check
> after a bunch of recursive calls that may change its value. As a result,
> it leads to stack-buffer-overflow during the instruction narrowing. This
> patch adds a missing check.
>
> Sergey Kaplun:
> * added the description and the test for the problem
>
> Part of tarantool/tarantool#10199
> ---
>
> Branch:https://github.com/tarantool/luajit/tree/skaplun/lj-1262-fix-limit-narrow-conv-backprop
> Related issues:
> *https://github.com/tarantool/tarantool/issues/10199
> *https://github.com/LuaJIT/LuaJIT/issues/1262
>
> src/lj_opt_narrow.c | 3 +-
> ...62-fix-limit-narrow-conv-backprop.test.lua | 61 +++++++++++++++++++
> 2 files changed, 63 insertions(+), 1 deletion(-)
> create mode 100644 test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
>
> diff --git a/src/lj_opt_narrow.c b/src/lj_opt_narrow.c
> index db0da10f..6b6f20d3 100644
> --- a/src/lj_opt_narrow.c
> +++ b/src/lj_opt_narrow.c
> @@ -341,7 +341,8 @@ static int narrow_conv_backprop(NarrowConv *nc, IRRef ref, int depth)
> NarrowIns *savesp = nc->sp;
> int count = narrow_conv_backprop(nc, ir->op1, depth);
> count += narrow_conv_backprop(nc, ir->op2, depth);
> - if (count <= 1) { /* Limit total number of conversions. */
> + /* Limit total number of conversions. */
> + if (count <= 1 && nc->sp < nc->maxsp) {
> *nc->sp++ = NARROWINS(IRT(ir->o, nc->t), ref);
> return count;
> }
> diff --git a/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
> new file mode 100644
> index 00000000..6bb4025d
> --- /dev/null
> +++ b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
> @@ -0,0 +1,61 @@
> +local tap = require('tap')
> +
> +-- Test file to demonstrate stack-buffer-overflow during the
> +-- narrowing optimization.
> +-- See also:https://github.com/LuaJIT/LuaJIT/issues/1262.
> +
> +local test = tap.test('lj-1262-fix-limit-narrow-conv-backprop'):skipcond({
> + ['Test requires JIT enabled'] = not jit.status(),
> +})
> +
> +test:plan(1)
> +
> +-- XXX: Test fails only under ASAN.
> +-- XXX: The original reproducer was found by fuzzer:
> +--https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=70779.
> +-- It creates a long side trace with a huge amount of ADD IRs,
> +-- which are recursively used in the `narrow_conv_backprop()` many
> +-- enough times to catch the stack-buffer-overflow. I can't
> +-- simplify the reproducer any more (or write it from scratch), so
> +-- I leave it in that state.
> +
I would remove a sentence "I can't simplify the reproducer ..."
from the test, because it is useful information on review and
I think it is useless for those who will use the test.
Feel free to ignore and leave as is.
s/any more/anymore/
> +local DEFAULT_NUMBER = 1
> +local tonumber = tonumber
> +
> +local always_number = function(val)
> + return tonumber(val) or DEFAULT_NUMBER
> +end
> +
> +local add = function(v1, v2)
> + return always_number(v1) + always_number(v2)
> +end
> +
> +jit.opt.start('hotloop=1', 'hotexit=1')
> +
> +local counter_0 = 0
> +local counter_1 = 0
> +local counter_2 = 0
> +local tmp = add(nil, 'Name')
> +local Name0 = add(tmp, 'Name')
> +-- Start a long side trace here.
> +for _ = 0, 0, 0 do
> + if counter_0 > 5 then break end
> + counter_0 = counter_0 + 1
> +
> + for _ = always_number(false), 1, always_number(Name0) do
> + if counter_1 > 5 then break end
> + counter_1 = counter_1 + 1
> +
> + repeat
> + if counter_2 > 5 then break end
> + counter_2 = counter_2 + 1
> +
> + Name0 = Name0 + Name0 + Name0
> + Name0 = add(Name0, nil) + Name0
> + until nil
> + end
> +end
> +
> +test:ok(true, 'no stack-buffer-overflow during narrowing')
> +
> +test:done(true)
[-- Attachment #2: Type: text/html, Size: 5192 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
@ 2024-09-17 12:01 ` Sergey Kaplun via Tarantool-patches
0 siblings, 0 replies; 6+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2024-09-17 12:01 UTC (permalink / raw)
To: Sergey Bronnikov; +Cc: tarantool-patches
Hi, Sergey!
Thanks for the review!
On 17.09.24, Sergey Bronnikov wrote:
> Hi, Sergey,
>
> thanks for tha patch! LGTM with a minor comment below.
>
> On 26.08.2024 15:37, Sergey Kaplun wrote:
> > From: Mike Pall <mike>
<snipped>
> > +-- XXX: Test fails only under ASAN.
> > +-- XXX: The original reproducer was found by fuzzer:
> > +--https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=70779.
> > +-- It creates a long side trace with a huge amount of ADD IRs,
> > +-- which are recursively used in the `narrow_conv_backprop()` many
> > +-- enough times to catch the stack-buffer-overflow. I can't
> > +-- simplify the reproducer any more (or write it from scratch), so
> > +-- I leave it in that state.
> > +
>
> I would remove a sentence "I can't simplify the reproducer ..."
>
> from the test, because it is useful information on review and
>
> I think it is useless for those who will use the test.
I do not agree -- it is kind of worning to prevent changing the test.
Also, we have something similar for gh-4199-gc64-fuse, for example.
>
> Feel free to ignore and leave as is.
So, ignoring for now.
>
>
> s/any more/anymore/
I decided to drop any, so the comment is the following now:
===================================================================
diff --git a/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
index 6bb4025d..c225fa14 100644
--- a/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
+++ b/test/tarantool-tests/lj-1262-fix-limit-narrow-conv-backprop.test.lua
@@ -15,9 +15,9 @@ test:plan(1)
-- https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=70779.
-- It creates a long side trace with a huge amount of ADD IRs,
-- which are recursively used in the `narrow_conv_backprop()` many
--- enough times to catch the stack-buffer-overflow. I can't
--- simplify the reproducer any more (or write it from scratch), so
--- I leave it in that state.
+-- enough times to catch the stack-buffer-overflow. I can't more
+-- simplify the reproducer any (or write it from scratch), so I
+-- leave it in that state.
local DEFAULT_NUMBER = 1
local tonumber = tonumber
===================================================================
<snipped>
--
Best regards,
Sergey Kaplun
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
2024-08-26 12:37 [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop() Sergey Kaplun via Tarantool-patches
2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
@ 2024-09-23 6:41 ` Maxim Kokryashkin via Tarantool-patches
2024-09-23 6:57 ` Sergey Kaplun via Tarantool-patches
2024-10-18 15:15 ` Sergey Kaplun via Tarantool-patches
2 siblings, 1 reply; 6+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2024-09-23 6:41 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
Hi, Sergey!
Thanks for the patch!
LGTM, except for a few nits below.
On Mon, Aug 26, 2024 at 03:37:40PM GMT, Sergey Kaplun wrote:
> From: Mike Pall <mike>
>
> Thanks to Sergey Kaplun.
>
> (cherry picked from commit e45fd4cb713b610506213692f3b55a1869febb03)
>
> `narrow_conv_backprop()` misses the stack pointer (`nc->sp`) limit check
> after a bunch of recursive calls that may change its value. As a result,
It took me a minute to understand that sentence. Let's paraphrase it the
following way:
| The function narrow_conv_backprop() does not include a limit
| check for the stack pointer (nc->sp), even though its value may
| change after several recursive calls.
> it leads to stack-buffer-overflow during the instruction narrowing. This
> patch adds a missing check.
Typo: s/a missing/the missing
>
> Sergey Kaplun:
> * added the description and the test for the problem
>
> Part of tarantool/tarantool#10199
> ---
>
> Branch: https://github.com/tarantool/luajit/tree/skaplun/lj-1262-fix-limit-narrow-conv-backprop
> Related issues:
> * https://github.com/tarantool/tarantool/issues/10199
> * https://github.com/LuaJIT/LuaJIT/issues/1262
<snipped>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
2024-09-23 6:41 ` Maxim Kokryashkin via Tarantool-patches
@ 2024-09-23 6:57 ` Sergey Kaplun via Tarantool-patches
0 siblings, 0 replies; 6+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2024-09-23 6:57 UTC (permalink / raw)
To: Maxim Kokryashkin; +Cc: tarantool-patches
Hi, Maxim!
Thanks for the review!
Fixed your comments and force-pushed the branch.
On 23.09.24, Maxim Kokryashkin wrote:
> Hi, Sergey!
> Thanks for the patch!
> LGTM, except for a few nits below.
>
> On Mon, Aug 26, 2024 at 03:37:40PM GMT, Sergey Kaplun wrote:
> > From: Mike Pall <mike>
> >
> > Thanks to Sergey Kaplun.
> >
> > (cherry picked from commit e45fd4cb713b610506213692f3b55a1869febb03)
> >
> > `narrow_conv_backprop()` misses the stack pointer (`nc->sp`) limit check
> > after a bunch of recursive calls that may change its value. As a result,
> It took me a minute to understand that sentence. Let's paraphrase it the
> following way:
>
> | The function narrow_conv_backprop() does not include a limit
> | check for the stack pointer (nc->sp), even though its value may
> | change after several recursive calls.
Rephrased as you suggested. The new commit message is the following:
| Fix limit check in narrow_conv_backprop().
|
| Thanks to Sergey Kaplun.
|
| (cherry picked from commit e45fd4cb713b610506213692f3b55a1869febb03)
|
| The function `narrow_conv_backprop()` does not include a limit check for
| the stack pointer (`nc->sp`), even though its value may change after
| several recursive calls. As a result, it leads to stack-buffer-overflow
| during the instruction narrowing. This patch adds the missing check.
|
| Sergey Kaplun:
| * added the description and the test for the problem
|
| Part of tarantool/tarantool#10199
>
> > it leads to stack-buffer-overflow during the instruction narrowing. This
> > patch adds a missing check.
> Typo: s/a missing/the missing
Fixed, thanks!
>
> >
> > Sergey Kaplun:
> > * added the description and the test for the problem
> >
> > Part of tarantool/tarantool#10199
> > ---
> >
> > Branch: https://github.com/tarantool/luajit/tree/skaplun/lj-1262-fix-limit-narrow-conv-backprop
> > Related issues:
> > * https://github.com/tarantool/tarantool/issues/10199
> > * https://github.com/LuaJIT/LuaJIT/issues/1262
>
> <snipped>
--
Best regards,
Sergey Kaplun
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop().
2024-08-26 12:37 [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop() Sergey Kaplun via Tarantool-patches
2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
2024-09-23 6:41 ` Maxim Kokryashkin via Tarantool-patches
@ 2024-10-18 15:15 ` Sergey Kaplun via Tarantool-patches
2 siblings, 0 replies; 6+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2024-10-18 15:15 UTC (permalink / raw)
To: Maxim Kokryashkin, Sergey Bronnikov; +Cc: tarantool-patches
I've applied the patch into all long-term branches in tarantool/luajit
and bumped a new version in master [1], release/3.2 [2] and
release/2.11 [3].
[1]: https://github.com/tarantool/tarantool/pull/10712
[2]: https://github.com/tarantool/tarantool/pull/10713
[3]: https://github.com/tarantool/tarantool/pull/10714
--
Best regards,
Sergey Kaplun
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2024-10-18 15:15 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-08-26 12:37 [Tarantool-patches] [PATCH luajit] Fix limit check in narrow_conv_backprop() Sergey Kaplun via Tarantool-patches
2024-09-17 10:32 ` Sergey Bronnikov via Tarantool-patches
2024-09-17 12:01 ` Sergey Kaplun via Tarantool-patches
2024-09-23 6:41 ` Maxim Kokryashkin via Tarantool-patches
2024-09-23 6:57 ` Sergey Kaplun via Tarantool-patches
2024-10-18 15:15 ` Sergey Kaplun via Tarantool-patches
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox