32/64 bit problem with myisam_max_sort_file_size

[Brian Aker]

tests/variables.test is failing on 32 bit for this:

--- ../tests/r/variables.result 2008-12-24 05:58:24.000000000 +0300
+++ ../tests/r/variables.reject 2008-12-24 06:12:07.000000000 +0300
@@ -452,11 +452,11 @@
 set global myisam_max_sort_file_size=default;
 select @@global.myisam_max_sort_file_size,@@local.max_join_size;
 @@global.myisam_max_sort_file_size @@local.max_join_size
-9223372036853727232 100
[phone]
 set @svc=@@global.myisam_max_sort_file_size, @svj=@@local.max_join_size;
 select @@global.myisam_max_sort_file_size,@@local.max_join_size;
 @@global.myisam_max_sort_file_size @@local.max_join_size
-9223372036853727232 100
[phone]
 set @@global.myisam_max_sort_file_size=111,@@local.max_join_size=222;
 select @@global.myisam_max_sort_file_size,@@local.max_join_size;
 @@global.myisam_max_sort_file_size @@local.max_join_size
@@ -470,7 +470,7 @@
 set @@global.myisam_max_sort_file_size=@svc, @@local.max_join_size=@svj;
 select @@global.myisam_max_sort_file_size,@@local.max_join_size;
 @@global.myisam_max_sort_file_size @@local.max_join_size
-9223372036853727232 100
[phone]
 set @a=1, @b=2;
 set @a=@b, @b=@a;
 select @a, @b;

[Monty Taylor (mordred)]

Yup. Joy of using diff for this. :)

Jay... does the new test framework allow us to test this more sensibly?
There are several variables that will have different max values on 32
and 64 bit like this. (of course, in _this_ case, I'm not sure why
myisam_max_sort_file_size is typed this way, since all of our file-size
variables should be 64-bit on either platform.) But the problem will be
there for memory buffers typed as size_t.

Monty

Brian Aker wrote:
> Public bug reported:
>
> tests/variables.test is failing on 32 bit for this:
>
> --- ../tests/r/variables.result 2008-12-24 05:58:24.000000000 +0300
> +++ ../tests/r/variables.reject 2008-12-24 06:12:07.000000000 +0300
> @@ -452,11 +452,11 @@
> set global myisam_max_sort_file_size=default;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> -9223372036853727232 100
> [phone]
> set @svc=@@global.myisam_max_sort_file_size, @svj=@@local.max_join_size;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> -9223372036853727232 100
> [phone]
> set @@global.myisam_max_sort_file_size=111,@@local.max_join_size=222;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> @@ -470,7 +470,7 @@
> set @@global.myisam_max_sort_file_size=@svc, @@local.max_join_size=@svj;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> -9223372036853727232 100
> [phone]
> set @a=1, @b=2;
> set @a=@b, @b=@a;
> select @a, @b;
>
> ** Affects: drizzle
> Importance: Undecided
> Status: New
>

[Brian Aker (brianaker)]

Size_t and time_t

Cheers,
    --Brian

On Dec 23, 2008, at 9:09 PM, Monty Taylor <email address hidden> wrote:

> Yup. Joy of using diff for this. :)
>
> Jay... does the new test framework allow us to test this more
> sensibly?
> There are several variables that will have different max values on 32
> and 64 bit like this. (of course, in _this_ case, I'm not sure why
> myisam_max_sort_file_size is typed this way, since all of our file-
> size
> variables should be 64-bit on either platform.) But the problem
> will be
> there for memory buffers typed as size_t.
>
> Monty
>
> Brian Aker wrote:
>> Public bug reported:
>>
>> tests/variables.test is failing on 32 bit for this:
>>
>> --- ../tests/r/variables.result 2008-12-24 05:58:24.000000000
>> +0300
>> +++ ../tests/r/variables.reject 2008-12-24 06:12:07.000000000
>> +0300
>> @@ -452,11 +452,11 @@
>> set global myisam_max_sort_file_size=default;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone] 100
>> set @svc=@@global.myisam_max_sort_file_size,
>> @svj=@@local.max_join_size;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone] 100
>> set @@global.myisam_max_sort_file_size=111,@@local.max_join_size=222;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> @@ -470,7 +470,7 @@
>> set @@global.myisam_max_sort_file_size=@svc,
>> @@local.max_join_size=@svj;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone] 100
>> set @a=1, @b=2;
>> set @a=@b, @b=@a;
>> select @a, @b;
>>
>> ** Affects: drizzle
>> Importance: Undecided
>> Status: New
>>
>
> --
> 32/64 bit problem with myisam_max_sort_file_size
> https://bugs.launchpad.net/bugs/311084
> You received this bug notification because you are a member of
> Drizzle-
> developers, which is subscribed to Drizzle.
>
> Status in A Lightweight SQL Database for Cloud and Web: New
>
> Bug description:
> tests/variables.test is failing on 32 bit for this:
>
> --- ../tests/r/variables.result 2008-12-24 05:58:24.000000000 +0300
> +++ ../tests/r/variables.reject 2008-12-24 06:12:07.000000000 +0300
> @@ -452,11 +452,11 @@
> set global myisam_max_sort_file_size=default;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> -9223372036853727232 100
> [phone] 100
> set @svc=@@global.myisam_max_sort_file_size,
> @svj=@@local.max_join_size;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_size @@local.max_join_size
> -9223372036853727232 100
> [phone] 100
> set @@global.myisam_max_sort_file_size=111,@@local.max_join_size=222;
> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
> @@global.myisam_max_sort_file_si...

[Jay Pipes (jaypipes)]

Monty Taylor wrote:
> Yup. Joy of using diff for this. :)
>
> Jay... does the new test framework allow us to test this more sensibly?
> There are several variables that will have different max values on 32
> and 64 bit like this. (of course, in _this_ case, I'm not sure why
> myisam_max_sort_file_size is typed this way, since all of our file-size
> variables should be 64-bit on either platform.) But the problem will be
> there for memory buffers typed as size_t.

Yes it will.

-jay

> Monty
>
> Brian Aker wrote:
>> Public bug reported:
>>
>> tests/variables.test is failing on 32 bit for this:
>>
>> --- ../tests/r/variables.result 2008-12-24 05:58:24.000000000 +0300
>> +++ ../tests/r/variables.reject 2008-12-24 06:12:07.000000000 +0300
>> @@ -452,11 +452,11 @@
>> set global myisam_max_sort_file_size=default;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone]
>> set @svc=@@global.myisam_max_sort_file_size, @svj=@@local.max_join_size;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone]
>> set @@global.myisam_max_sort_file_size=111,@@local.max_join_size=222;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> @@ -470,7 +470,7 @@
>> set @@global.myisam_max_sort_file_size=@svc, @@local.max_join_size=@svj;
>> select @@global.myisam_max_sort_file_size,@@local.max_join_size;
>> @@global.myisam_max_sort_file_size @@local.max_join_size
>> -9223372036853727232 100
>> [phone]
>> set @a=1, @b=2;
>> set @a=@b, @b=@a;
>> select @a, @b;
>>
>> ** Affects: drizzle
>> Importance: Undecided
>> Status: New
>>
>

[Lee Bieber (kalebral-deactivatedaccount)]

This is no longer issue, possibly Stewart fixed this in his slew of Solaris fixes.