Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-10 Thread Michael Paquier
On Mon, Sep 09, 2019 at 09:43:06AM +0900, Michael Paquier wrote: > On Sun, Sep 08, 2019 at 07:44:16PM -0400, Tom Lane wrote: >> The whole thing should be back-patched into branches that have >> any affected tests. (But, please, not till after beta4 is >> tagged.) > > Sure. Don't worry about that

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Michael Paquier
On Sun, Sep 08, 2019 at 07:44:16PM -0400, Tom Lane wrote: > Yeah, perhaps, but I wasn't sure about how to do that easily. > Feel free to add it ... Feeding the makefile contents and then doing a lookup using =~ should be enough. I think that we should just refactor set of routines for fetchTests(

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Andrew Dunstan
On 9/8/19 6:18 PM, Tom Lane wrote: > Andrew Dunstan writes: >> On 9/8/19 5:59 PM, Tom Lane wrote: >>> Hm. Changing the buildfarm script would be an alternative way to >>> fix the issue so far as the buildfarm is concerned, but it doesn't >>> seem like it provides any useful way for one to manua

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Tom Lane
Michael Paquier writes: > I think that it is a good idea to add in subdircheck an extra path to > check after TAP tests and run optionally these on top of the normal > regression tests. I have a couple of comments. > Shouldn't we check after TAP_TESTS in the Makefile? Yeah, perhaps, but I wasn'

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Michael Paquier
On Sun, Sep 08, 2019 at 06:18:33PM -0400, Tom Lane wrote: > Sure, I saw that you can run one test that way ... but what do you > do when you want the equivalent of check-world? I think that it is a good idea to add in subdircheck an extra path to check after TAP tests and run optionally these on t

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Tom Lane
Andrew Dunstan writes: > On 9/8/19 5:59 PM, Tom Lane wrote: >> Hm. Changing the buildfarm script would be an alternative way to >> fix the issue so far as the buildfarm is concerned, but it doesn't >> seem like it provides any useful way for one to manually invoke >> the tests on Windows. Or am

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Andrew Dunstan
On 9/8/19 5:59 PM, Tom Lane wrote: > Andrew Dunstan writes: >> On 9/8/19 12:07 PM, Tom Lane wrote: >>> It looks to me like the reason is that src/tools/msvc/vcregress.pl's >>> subroutine subdircheck isn't considering the possibility that >>> subdirectories of src/test/modules contain TAP tests.

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Tom Lane
Andrew Dunstan writes: > On 9/8/19 12:07 PM, Tom Lane wrote: >> It looks to me like the reason is that src/tools/msvc/vcregress.pl's >> subroutine subdircheck isn't considering the possibility that >> subdirectories of src/test/modules contain TAP tests. The >> same code is used for contrib, so s

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Andrew Dunstan
On 9/8/19 12:07 PM, Tom Lane wrote: > I noticed $subject while checking to see if commit db4383189's > new test script was behaving properly in the buildfarm. dory, > for one, should be running it but it just isn't. > > It looks to me like the reason is that src/tools/msvc/vcregress.pl's > subro

Re: MSVC buildfarm critters are not running modules' TAP tests

2019-09-08 Thread Christoph Moench-Tegeder
## Tom Lane (t...@sss.pgh.pa.us): > I took a stab at fixing this, but lacking a Windows environment > to test in, I can't be sure if it works. The attached does kinda > sorta work if I run it in a Linux environment --- but I found that > system() doesn't automatically expand "t/*.pl" on Linux. I