Thread: pg_regress writes into source tree
When using a vpath build pg_regress writes the processed input/*.source files into the *source* tree, which isn't supposed to happen. This appears to be a thinko introduced in this patch: e3fc4a97bc8ee82a78605b5ffe79bd4cf3c6213b The attached patch fixes it.
Attachment
Peter Eisentraut wrote: > When using a vpath build pg_regress writes the processed input/*.source > files into the *source* tree, which isn't supposed to happen. > > This appears to be a thinko introduced in this patch: > e3fc4a97bc8ee82a78605b5ffe79bd4cf3c6213b Oh, I noticed this while doing the dummy_seclabel move to src/test/modules and I thought it was on purpose; if I'm not mistaken this is why we had to add the .sql file to .gitignore. Another thing in that patch was that I had to add the sql/ directory to the source tree, but other than that .gitignore file it was empty. Maybe pg_regress should create the sql/ directory in the build dir if it doesn't exist. This is only a problem if a pg_regress suite only runs stuff from input/, because otherwise the sql/ dir already exists in the source. -- Álvaro Herrera http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
Hi, On 2014-12-11 22:02:26 -0500, Peter Eisentraut wrote: > When using a vpath build pg_regress writes the processed input/*.source > files into the *source* tree, which isn't supposed to happen. > > This appears to be a thinko introduced in this patch: > e3fc4a97bc8ee82a78605b5ffe79bd4cf3c6213b > > The attached patch fixes it. I've been annoyed by this more than once, specifically when trying to run tests with different compilation settings at once... Greetings, Andres Freund -- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services
On Fri, Dec 12, 2014 at 10:45 PM, Alvaro Herrera <alvherre@2ndquadrant.com> wrote: > Another thing in that patch was that I had to add the sql/ directory to > the source tree, but other than that .gitignore file it was empty. > Maybe pg_regress should create the sql/ directory in the build dir if it > doesn't exist. This is only a problem if a pg_regress suite only runs > stuff from input/, because otherwise the sql/ dir already exists in the > source. +1 for having pg_regress create the sql/ directory when it does not exist. Current behavior is annoying when modules having only tests in input/... -- Michael
On 12/18/2014 03:02 AM, Michael Paquier wrote: > On Fri, Dec 12, 2014 at 10:45 PM, Alvaro Herrera > <alvherre@2ndquadrant.com> wrote: > >> Another thing in that patch was that I had to add the sql/ directory to >> the source tree, but other than that .gitignore file it was empty. >> Maybe pg_regress should create the sql/ directory in the build dir if it >> doesn't exist. This is only a problem if a pg_regress suite only runs >> stuff from input/, because otherwise the sql/ dir already exists in the >> source. > +1 for having pg_regress create the sql/ directory when it does not > exist. Current behavior is annoying when modules having only tests in > input/... That seems like a separate issue. I think Peter should commit his patch and backpatch it immediately, and we can deal with the missing sql directory when someone sends in a patch. cheers andrew
On 12/18/2014 06:05 PM, Andrew Dunstan wrote: > > On 12/18/2014 03:02 AM, Michael Paquier wrote: >> On Fri, Dec 12, 2014 at 10:45 PM, Alvaro Herrera >> <alvherre@2ndquadrant.com> wrote: >> >>> Another thing in that patch was that I had to add the sql/ directory to >>> the source tree, but other than that .gitignore file it was empty. >>> Maybe pg_regress should create the sql/ directory in the build dir >>> if it >>> doesn't exist. This is only a problem if a pg_regress suite only runs >>> stuff from input/, because otherwise the sql/ dir already exists in the >>> source. >> +1 for having pg_regress create the sql/ directory when it does not >> exist. Current behavior is annoying when modules having only tests in >> input/... > > > > That seems like a separate issue. I think Peter should commit his > patch and backpatch it immediately, and we can deal with the missing > sql directory when someone sends in a patch. What's happened on this? cheers andrew
On Sat, Feb 14, 2015 at 6:24 AM, Andrew Dunstan <andrew@dunslane.net> wrote: > > On 12/18/2014 06:05 PM, Andrew Dunstan wrote: >> >> >> On 12/18/2014 03:02 AM, Michael Paquier wrote: >>> >>> On Fri, Dec 12, 2014 at 10:45 PM, Alvaro Herrera >>> <alvherre@2ndquadrant.com> wrote: >>> >>>> Another thing in that patch was that I had to add the sql/ directory to >>>> the source tree, but other than that .gitignore file it was empty. >>>> Maybe pg_regress should create the sql/ directory in the build dir if it >>>> doesn't exist. This is only a problem if a pg_regress suite only runs >>>> stuff from input/, because otherwise the sql/ dir already exists in the >>>> source. >>> >>> +1 for having pg_regress create the sql/ directory when it does not >>> exist. Current behavior is annoying when modules having only tests in >>> input/... >> >> >> >> >> That seems like a separate issue. I think Peter should commit his patch >> and backpatch it immediately, and we can deal with the missing sql directory >> when someone sends in a patch. > > > > What's happened on this? Nothing has been committed, and as far as I understood this patch could have been simply pushed... -- Michael
On 2/13/15 11:20 PM, Michael Paquier wrote: > On Sat, Feb 14, 2015 at 6:24 AM, Andrew Dunstan <andrew@dunslane.net> wrote: >> >> On 12/18/2014 06:05 PM, Andrew Dunstan wrote: >>> >>> >>> On 12/18/2014 03:02 AM, Michael Paquier wrote: >>>> >>>> On Fri, Dec 12, 2014 at 10:45 PM, Alvaro Herrera >>>> <alvherre@2ndquadrant.com> wrote: >>>> >>>>> Another thing in that patch was that I had to add the sql/ directory to >>>>> the source tree, but other than that .gitignore file it was empty. >>>>> Maybe pg_regress should create the sql/ directory in the build dir if it >>>>> doesn't exist. This is only a problem if a pg_regress suite only runs >>>>> stuff from input/, because otherwise the sql/ dir already exists in the >>>>> source. >>>> >>>> +1 for having pg_regress create the sql/ directory when it does not >>>> exist. Current behavior is annoying when modules having only tests in >>>> input/... >>> >>> >>> >>> >>> That seems like a separate issue. I think Peter should commit his patch >>> and backpatch it immediately, and we can deal with the missing sql directory >>> when someone sends in a patch. >> >> >> >> What's happened on this? > > Nothing has been committed, and as far as I understood this patch > could have been simply pushed... Just so this doesn't get lost... did something make it into a CommitFest on this? -- Jim Nasby, Data Architect, Blue Treble Consulting Data in Trouble? Get it in Treble! http://BlueTreble.com
On Wed, Apr 8, 2015 at 10:28 AM, Jim Nasby wrote: > Just so this doesn't get lost... did something make it into a CommitFest on > this? Peter's patch has been committed as 64cdbbc, while the idea to create sql/ by pg_regress if it is not present did not gather much interest in this CF: https://commitfest.postgresql.org/4/100/ -- Michael