On 11/19/2012 02:36 PM, Stefano Lattarini wrote: > On 11/19/2012 12:21 PM, Théophile Ranquet wrote: >> Hi Stefano, >> >>> From latest master: >>> >>> $ make check >>> GEN public-submodule-commit >>> fatal: Not a valid object name origin >>> /usr/local/bleeding/libexec/git-core/git-submodule: line 445: \ >>> test: daf7f8c02242c535d596231e2f655109b97fa2bc: unary operator expected >>> Stopping at 'gnulib'; script returned non-zero status. >>> maint.mk: found non-public submodule commit >>> maint.mk:1391: recipe for target 'public-submodule-commit' failed >>> make: *** [public-submodule-commit] Error 1 >>> >>> I haven't investigated any further. Can you reproduce the issue? >> >> This might be related to your Git, which version are you using ? >> > $ git --version > git version 1.8.0.150.gb0b00a3 > >> I have failed to reproduce with 1.7.11.4 and 1.8.0. >> > Could this be a git regression then? I'll try updating to latest Git > master, and see if the error disappears. > No, the error is still there (with git 1.8.0.209.gf3828dc).
But then, I've also verified that the error is present with Git 1.7.10.4 (installed through Debian packages): $ PATH=/usr/bin:$PATH make public-submodule-commit GEN public-submodule-commit fatal: Not a valid object name origin /usr/lib/git-core/git-submodule: line 354: test: daf7f8c02242c535d596231e2f655109b97fa2bc: unary operator expected Stopping at 'gnulib'; script returned non-zero status. maint.mk: found non-public submodule commit make: *** [public-submodule-commit] Error 1 In addition, the 'public-submodule-commit' recipe runs successfully for other GNU packages (like cppi), even if I use the bleeding-edge Git. So the problem must actually be in Bison. Regards, Stefano
