Hi,
On 2021-05-09 15:57:22 +0300, Yura Sokolov wrote:
> Occasionally there is a need to run cassert builds in production to
> catch an issue. It is usually ok if cassert build O(1) slower than
> optimized biuld (ie it is slower in some constant factor C). But
> if cassert build will be quadratical
David Rowley писал 2021-05-09 04:01:
On Sun, 9 May 2021 at 03:29, Pavel Stehule
wrote:
Personally, I have not problem with too slow assertions, although it
is not too practical. The main problem is some shock, and feeling so
some is wrong. I spent 1 hour detecting if it is a bug or not.
Than
ne 9. 5. 2021 v 3:01 odesílatel David Rowley napsal:
> On Sun, 9 May 2021 at 03:29, Pavel Stehule
> wrote:
> > Personally, I have not problem with too slow assertions, although it is
> not too practical. The main problem is some shock, and feeling so some is
> wrong. I spent 1 hour detecting if
On Sun, 9 May 2021 at 03:29, Pavel Stehule wrote:
> Personally, I have not problem with too slow assertions, although it is not
> too practical. The main problem is some shock, and feeling so some is wrong.
> I spent 1 hour detecting if it is a bug or not.
Thanks for spending the time figuring
On Sun, 9 May 2021 at 00:07, Tomas Vondra wrote:
>
>
> On 5/8/21 1:27 PM, David Rowley wrote:
> > On Sat, 8 May 2021 at 22:33, Tomas Vondra
> > wrote:
> >> I don't know if there's a better way to do these tests, but if there's
> >> not I'd not worry about it too much for now.
> >
> > So you're -
Pavel Stehule writes:
> so 8. 5. 2021 v 9:39 odesílatel David Rowley napsal:
>> So, I'm -1 for "heavy asserts".
> Personally, I have not problem with too slow assertions, although it is not
> too practical.
I'm very much on David's side here. I'm currently trying to figure out
why CLOBBER_CACH
so 8. 5. 2021 v 9:39 odesílatel David Rowley napsal:
> On Sat, 8 May 2021 at 19:03, Yura Sokolov
> wrote:
> > Perhaps there is need for flag "heavy asserts". Or option
> "--enable-cassert=heavy". Then USE_ASSERT_CHECKING could be defined to
> integer 1 or 2 depending on "heaviness" of enabled ch
On 5/8/21 1:27 PM, David Rowley wrote:
On Sat, 8 May 2021 at 22:33, Tomas Vondra wrote:
I don't know if there's a better way to do these tests, but if there's
not I'd not worry about it too much for now.
So you're -1 on the proposed patch?
Oh! I have not noticed there was a patch. No, I'
On Sat, 8 May 2021 at 22:33, Tomas Vondra wrote:
> I don't know if there's a better way to do these tests, but if there's
> not I'd not worry about it too much for now.
So you're -1 on the proposed patch?
David
On 5/8/21 9:39 AM, David Rowley wrote:
On Sat, 8 May 2021 at 19:03, Yura Sokolov wrote:
Perhaps there is need for flag "heavy asserts". Or option "--enable-cassert=heavy". Then
USE_ASSERT_CHECKING could be defined to integer 1 or 2 depending on "heaviness" of enabled checks.
I'd rather ex
On Sat, 8 May 2021 at 19:03, Yura Sokolov wrote:
> Perhaps there is need for flag "heavy asserts". Or option
> "--enable-cassert=heavy". Then USE_ASSERT_CHECKING could be defined to
> integer 1 or 2 depending on "heaviness" of enabled checks.
I'd rather explore all other options before we went
8 мая 2021 г. 00:16:54 GMT+03:00, Tomas Vondra
пишет:
>On 5/7/21 11:04 PM, David Rowley wrote:
>> On Sat, 8 May 2021 at 08:18, Pavel Stehule
>wrote:
>>>
>>> pá 7. 5. 2021 v 21:56 odesílatel David Rowley
>napsal:
With USE_ASSERT_CHECKING builds, I did add some code that verifies
>the
>>>
On Sat, 8 May 2021 at 14:43, Justin Pryzby wrote:
> You could put this into a separate function called by ExecEndResultCache().
> Then anyone that breaks the memory accounting can also call the function in
> the
> places they changed to help figure out what they broke.
I almost did it that way a
On Sat, May 08, 2021 at 02:26:44PM +1200, David Rowley wrote:
> On Sat, 8 May 2021 at 09:16, Tomas Vondra
> wrote:
> > On 5/7/21 11:04 PM, David Rowley wrote:
> > > Another thought I have is that maybe it would be ok just to move
> > > memory accounting debug code so it only runs once in
> > > Ex
On Sat, 8 May 2021 at 09:16, Tomas Vondra wrote:
>
> On 5/7/21 11:04 PM, David Rowley wrote:
> > Another thought I have is that maybe it would be ok just to move
> > memory accounting debug code so it only runs once in
> > ExecEndResultCache. I struggling to imagine that if the memory
> > trackin
On 5/7/21 11:04 PM, David Rowley wrote:
On Sat, 8 May 2021 at 08:18, Pavel Stehule wrote:
pá 7. 5. 2021 v 21:56 odesílatel David Rowley napsal:
With USE_ASSERT_CHECKING builds, I did add some code that verifies the
memory tracking is set correctly when evicting from the cache. This
code is p
On Sat, 8 May 2021 at 08:18, Pavel Stehule wrote:
>
> pá 7. 5. 2021 v 21:56 odesílatel David Rowley napsal:
>> With USE_ASSERT_CHECKING builds, I did add some code that verifies the
>> memory tracking is set correctly when evicting from the cache. This
>> code is pretty expensive as it loops over
pá 7. 5. 2021 v 21:56 odesílatel David Rowley napsal:
> On Sat, 8 May 2021 at 07:18, Pavel Stehule
> wrote:
> > yes, the slowdown is related to debug assertions
>
> With USE_ASSERT_CHECKING builds, I did add some code that verifies the
> memory tracking is set correctly when evicting from the ca
On Sat, 8 May 2021 at 07:18, Pavel Stehule wrote:
> yes, the slowdown is related to debug assertions
With USE_ASSERT_CHECKING builds, I did add some code that verifies the
memory tracking is set correctly when evicting from the cache. This
code is pretty expensive as it loops over the entire cach
pá 7. 5. 2021 v 21:06 odesílatel Yura Sokolov
napsal:
> Pavel Stehule писал 2021-05-07 21:45:
> >>
> >> Samples: 119K of event 'cycles', 4000 Hz, Event count (approx.):
> >> Overhead Shared Object Symbol
> >> 79.20% postgres [.] cache_reduce_memory
> >> 1.94% [kern
On 5/7/21 9:06 PM, Yura Sokolov wrote:
Pavel Stehule писал 2021-05-07 21:45:
Samples: 119K of event 'cycles', 4000 Hz, Event count (approx.):
Overhead Shared Object Symbol
79.20% postgres [.] cache_reduce_memory
1.94% [kernel] [k] native_write
Pavel Stehule писал 2021-05-07 21:45:
Samples: 119K of event 'cycles', 4000 Hz, Event count (approx.):
Overhead Shared Object Symbol
79.20% postgres [.] cache_reduce_memory
1.94% [kernel] [k] native_write_msr_safe
1.63% [kernel]
pá 7. 5. 2021 v 20:24 odesílatel Pavel Stehule
napsal:
> Hi
>
> I am testing new features of Postgres 14, and now I am trying to check the
> result cache. Unfortunately on my test data, the result is not too good.
> the behaviour is very non linear. Is it expected?
>
> create table t1(a int, t2_i
23 matches
Mail list logo
