[tarantool-patches] Re: [PATCH v1 1/1] lua: fix strange behaviour of tonumber64

[Vladislav Shpilevoy]

On 16/07/2018 15:49, Alexander Turenko wrote:
> Hi, Vlad!
> 
> That is interesting discussion. Hope you don't mind my participation.

Hi! Your participation is appreciated!

> 
> WBR, Alexander Turenko.
> 
> On Mon, Jul 16, 2018 at 01:23:36PM +0300, Vladislav Shpilevoy wrote:
>> Thanks for the patch! See 4 comments below.
>>
>> On 13/07/2018 14:21, Kirill Shcherbatov wrote:
>>> Function tonumber64 has worked incorrectly with values less
>>> than LLONG_MIN.
>>> Now it works in the interval [LLONG_MIN, ULLONG_MAX] returning
>>> nil otherwise.
>>>
>>> Closes #3466.
>>> ---
>>> Branch: https://github.com/tarantool/tarantool/compare/kshch/gh-3466-tonumber64-strange-behaviour
>>> Issue: https://github.com/tarantool/tarantool/issues/3466
>>>
>>>    src/lua/init.c         |  6 +++++-
>>>    test/box/misc.result   | 19 +++++++++++++++++++
>>>    test/box/misc.test.lua |  8 ++++++++
>>>    3 files changed, 32 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/src/lua/init.c b/src/lua/init.c
>>> index 9a96030..4b5285d 100644
>>> --- a/src/lua/init.c
>>> +++ b/src/lua/init.c
>>> @@ -222,7 +222,11 @@ lbox_tonumber64(struct lua_State *L)
>>>    			if (argl == 0) {
>>>    				lua_pushnil(L);
>>>    			} else if (negative) {
>>> -				luaL_pushint64(L, -1 * (long long )result);
>>> +				if (result > -((unsigned long long )LLONG_MIN)) {
>>
>> 1. Please, do not enclose one-line bodies into {}.
>>
>> 2. How can you cast LLONG_MIN (that is negative) to the unsigned type?
>>
> 
> Cast does not change bits. It is legal.

Yes, technically it is legal, but casting negative value to an unsigned type
looks weird.

> 
>> 3. Why not 'result > LLONG_MAX'? As I understand, abs(LLONG_MAX) == abs(LLONG_MIN),
>> it is not? (http://www.cplusplus.com/reference/climits/)
>>
> 
> No, LLONG_MAX is 2^63-1, but LLONG_MIN is -2^63. We want to compare
> result with 2^63. We are trying to do so in platform-independent way
> (hovewer unsiged unary nimus equivalence with signed one is likely
> two-complement number representation property and can be violated on
> other platforms).
> 
> Are you think we should introduce our own constant
> 9223372036854775808ULL (2^63) and avoid that complex assumptions set? It

Ultimately no. We should not invent the constants.

> would be explicitly number-representation-dependent, so maybe it is
> better.

Ok. Logically we want an error on -result < INT64_MIN, right?
It is the same as result > -INT64_MIN. But we can not say
-INT64_MIN because abs(INT64_MIN) > INT64_MAX, yes?

Then lets rephrase the comparison:

     result > -INT64_MIN
            |
            v
   result + 1 >= -INT64_MIN
            |
            v
     result >= -INT64_MIN - 1
            |
            v
    result >= -(INT64_MIN + 1) <- that is the solution.

As I understand, -(INT64_MIN + 1) is exactly 2^63 - 1 and
fits in int64, right?