From: Sergey Bronnikov via Tarantool-patches <tarantool-patches@dev.tarantool.org>
To: Maxim Kokryashkin <m.kokryashkin@tarantool.org>,
Sergey Bronnikov <estetus@gmail.com>
Cc: max.kokryashkin@gmail.com, tarantool-patches@dev.tarantool.org
Subject: Re: [Tarantool-patches] [PATCH luajit 2/2][v2] Followup fix for embedded bytecode loader.
Date: Mon, 4 Sep 2023 19:34:01 +0300 [thread overview]
Message-ID: <59ea8e26-071f-1819-50c8-c63ee582505d@tarantool.org> (raw)
In-Reply-To: <yeiqzojpelto6wdubzk76gpkretkzruhnk2h45jbe34ljfevyq@6h4bpqrotzi2>
Hi, Max
On 9/1/23 13:05, Maxim Kokryashkin via Tarantool-patches wrote:
> On Thu, Aug 31, 2023 at 02:32:14PM +0300, Sergey Bronnikov via Tarantool-patches wrote:
>> From: Sergey Bronnikov <sergeyb@tarantool.org>
<snipped>
>> +
>> +/**
>> + * Function generates a huge chunk of "bytecode" with a size bigger than
>> + * LJ_MAX_BUF. Generated chunk must enable endmark in a Lex state.
>> + */
>> +static const char *
>> +bc_reader_with_endmark(lua_State *L, void *data, size_t *size)
>> +{
>> + UNUSED(data);
>> + int bc_chunk_size = (size_t)0;
>> + static char *bc_chunk = NULL;
>> + free(bc_chunk);
> What's the point of free here? Why the buffer is static?
Because callee (aka Reader) is responsible for buffer, reader
initializes it once
and then reuse.
> The block must exist until the reader function is called again.
> To signal the end of the chunk, the reader must return NULL or set
size to zero.
> The reader function may return pieces of any size greater than zero.
1. http://www.lua.org/manual/5.1/manual.html#lua_Reader
>> +
>> + bc_chunk = malloc(bc_chunk_size);
> Malloc of zero size doesn't seem to be the thing you wanted to do.
Right. Updated:
@@ -33,26 +33,9 @@ static const char *
bc_reader_with_endmark(lua_State *L, void *data, size_t *size)
{
UNUSED(data);
- int bc_chunk_size = (size_t)0;
- static char *bc_chunk = NULL;
- free(bc_chunk);
-
- bc_chunk = malloc(bc_chunk_size);
- assert(bc_chunk != NULL);
-
- /**
- * `lua_load` automatically detects whether the chunk is text or
binary,
- * and loads it accordingly. We need a trace for bytecode input,
- * so it is necessary to deceive a check in lj_lex_setup, that
- * makes a sanity check and detects whether input is bytecode or text
- * by the first char. Put LUA_SIGNATURE[0] at the beginning of the
- * allocated region.
- */
- bc_chunk[0] = LUA_SIGNATURE[0];
-
- *size = bc_chunk_size;
+ *size = ~(size_t)0;
- return bc_chunk;
+ return NULL;
}
static int bc_loader_with_endmark(void *test_state)
>> + assert(bc_chunk != NULL);
>> +
>> + /**
>> + * `lua_load` automatically detects whether the chunk is text or binary,
>> + * and loads it accordingly. We need a trace for bytecode input,
>> + * so it is necessary to deceive a check in lj_lex_setup, that
>> + * makes a sanity check and detects whether input is bytecode or text
>> + * by the first char. Put LUA_SIGNATURE[0] at the beginning of the
>> + * allocated region.
>> + */
>> + bc_chunk[0] = LUA_SIGNATURE[0];
>> +
>> + *size = bc_chunk_size;
>> +
>> + return bc_chunk;
>> +}
>> +
>> +static int bc_loader_with_endmark(void *test_state)
>> +{
>> + lua_State *L = test_state;
>> + void *ud = NULL;
>> + int res = lua_load(L, bc_reader_with_endmark, ud, "endmark");
>> +
>> + /*
>> + * Make sure we passed the condition with lj_err_mem in the function
>> + * `lex_more`.
>> + */
>> + assert_true(res != LUA_ERRMEM);
>> +
>> + return TEST_EXIT_SUCCESS;
>> +}
>> +
>> +enum bc_emission_state {
>> + EMIT_BC,
>> + EMIT_EOF,
>> +};
>> +
>> +typedef struct {
>> + enum bc_emission_state state;
>> +} dt;
>> +
>> +/**
>> + * Function returns a bytecode chunk on the first call and NULL and size equal
>> + * to zero on the second call. Triggers the END_OF_STREAM flag in the function
>> + * `lex_more`.
>> + */
>> +static const char *
>> +bc_reader_with_eof(lua_State *L, void *data, size_t *size)
>> +{
>> + UNUSED(data);
>> + UNUSED(L);
>> + dt *test_data = (dt *)data;
>> + if (test_data->state == EMIT_EOF) {
> This section is unreachable, isn't it?
Right, fixed it.
>> + *size = 0;
>> + return NULL;
>> + }
>> +
>> + static char *bc_chunk = NULL;
>> + free(bc_chunk);
> Ditto.
>> +
>> + size_t sz = 10;
> Is there any reason for it to be exactly 10? Drop a comment.
Set it to 2 and dropped a comment.
@@ -98,10 +81,23 @@ bc_reader_with_eof(lua_State *L, void *data, size_t
*size)
static char *bc_chunk = NULL;
free(bc_chunk);
- size_t sz = 10;
+ /**
+ * Minimal size of a buffer with bytecode:
+ * signiture (1 byte) and a bytecode itself (1 byte).
+ */
+ size_t sz = 2;
bc_chunk = malloc(sz);
+ /**
+ * `lua_load` automatically detects whether the chunk is text or
binary,
+ * and loads it accordingly. We need a trace for *bytecode* input,
+ * so it is necessary to deceive a check in `lj_lex_setup`, that
+ * makes a sanity check and detects whether input is bytecode or text
+ * by the first char. Put `LUA_SIGNATURE[0]` at the beginning of the
+ * allocated region.
+ */
bc_chunk[0] = LUA_SIGNATURE[0];
*size = sz;
+ test_data->state = EMIT_EOF;
return bc_chunk;
}
<snipped>
next prev parent reply other threads:[~2023-09-04 16:34 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-31 11:29 [Tarantool-patches] [PATCH luajit 0/2][v2] Fix " Sergey Bronnikov via Tarantool-patches
2023-08-31 11:30 ` [Tarantool-patches] [PATCH luajit 1/2][v2] " Sergey Bronnikov via Tarantool-patches
2023-08-31 11:49 ` Sergey Bronnikov via Tarantool-patches
2023-09-01 9:42 ` Maxim Kokryashkin via Tarantool-patches
2023-09-04 9:31 ` Sergey Bronnikov via Tarantool-patches
2023-09-05 6:34 ` Maxim Kokryashkin via Tarantool-patches
2023-09-05 14:10 ` Sergey Kaplun via Tarantool-patches
2023-09-07 15:21 ` Sergey Bronnikov via Tarantool-patches
2023-09-11 8:45 ` Sergey Kaplun via Tarantool-patches
2023-09-12 10:20 ` Sergey Bronnikov via Tarantool-patches
2023-10-31 11:30 ` Sergey Kaplun via Tarantool-patches
2023-09-05 14:12 ` Sergey Kaplun via Tarantool-patches
2023-09-07 7:06 ` Sergey Bronnikov via Tarantool-patches
2023-08-31 11:32 ` [Tarantool-patches] [PATCH luajit 2/2][v2] Followup fix for " Sergey Bronnikov via Tarantool-patches
2023-09-01 10:05 ` Maxim Kokryashkin via Tarantool-patches
2023-09-04 16:34 ` Sergey Bronnikov via Tarantool-patches [this message]
2023-09-05 6:45 ` Maxim Kokryashkin via Tarantool-patches
2023-09-05 12:55 ` Sergey Kaplun via Tarantool-patches
2023-09-07 7:04 ` Sergey Bronnikov via Tarantool-patches
2023-09-11 9:26 ` Sergey Kaplun via Tarantool-patches
2023-09-12 10:30 ` Sergey Bronnikov via Tarantool-patches
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=59ea8e26-071f-1819-50c8-c63ee582505d@tarantool.org \
--to=tarantool-patches@dev.tarantool.org \
--cc=estetus@gmail.com \
--cc=m.kokryashkin@tarantool.org \
--cc=max.kokryashkin@gmail.com \
--cc=sergeyb@tarantool.org \
--subject='Re: [Tarantool-patches] [PATCH luajit 2/2][v2] Followup fix for embedded bytecode loader.' \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox