On 4/1/21 4:24 AM, David Rowley wrote:
> On Thu, 1 Apr 2021 at 14:49, Tomas Vondra
> wrote:
>> I think that's probably sufficient for now. More testing may be useful
>> once I refactor the code a bit to make it more readable.
>
> Just to let you know, I left it running a bit longer. About 70 min
On Thu, 1 Apr 2021 at 14:49, Tomas Vondra wrote:
> I think that's probably sufficient for now. More testing may be useful
> once I refactor the code a bit to make it more readable.
Just to let you know, I left it running a bit longer. About 70 mins
now. No crashes yet.
David
On 4/1/21 3:43 AM, David Rowley wrote:
> On Thu, 1 Apr 2021 at 11:07, Tomas Vondra
> wrote:
>> The attached patch fixes this for me. David, can you check if this
>> resolves the issue for you?
>
> Thanks for taking a look at this.
>
> I applied your patch and kicked off sqlancer with 64 threads
On Thu, 1 Apr 2021 at 11:07, Tomas Vondra wrote:
> The attached patch fixes this for me. David, can you check if this
> resolves the issue for you?
Thanks for taking a look at this.
I applied your patch and kicked off sqlancer with 64 threads. It's
been going about 30 mins without any crashes so
On 3/31/21 7:54 PM, Tomas Vondra wrote:
> On 3/31/21 7:08 PM, Alvaro Herrera wrote:
>> On 2021-Mar-31, Tomas Vondra wrote:
>>
>>> Thanks for the report, I'll take a look. You're right this seems like an
>>> out-of-bounds access, but mcv_match_expression is only expected to be
>>> run on expressio
On 3/31/21 7:08 PM, Alvaro Herrera wrote:
> On 2021-Mar-31, Tomas Vondra wrote:
>
>> Thanks for the report, I'll take a look. You're right this seems like an
>> out-of-bounds access, but mcv_match_expression is only expected to be
>> run on expressions we know are in the statistics (because we pic
On 2021-Mar-31, Tomas Vondra wrote:
> Thanks for the report, I'll take a look. You're right this seems like an
> out-of-bounds access, but mcv_match_expression is only expected to be
> run on expressions we know are in the statistics (because we pick the
> statistics like that). Clearly, that does
On 3/31/21 10:05 AM, David Rowley wrote:
> Hi Tomas,
>
> I'm debugging a crash after running sqllancer on current master. The
> first bad commit seems to be this one.
>
> The crash stack trace is:
>
> Program received signal SIGSEGV, Segmentation fault.
> pg_detoast_datum_packed (datum=0x5) a
Hi Tomas,
I'm debugging a crash after running sqllancer on current master. The
first bad commit seems to be this one.
The crash stack trace is:
Program received signal SIGSEGV, Segmentation fault.
pg_detoast_datum_packed (datum=0x5) at fmgr.c:1759
1759 if (VARATT_IS_COMPRESSED(datum) || VARATT_I
