Custom Query (431 matches)
Results (19 - 21 of 431)
Ticket | Resolution | Summary | Owner | Reporter |
---|---|---|---|---|
#1436 | fixed | man jsv_script_interface shows wrong call in examples section | Dave Love <d.love@…> | Reuti |
Description |
The man page and html document contain at the end: function call returned value ------------------------------------------------------ jsv_is_param(l_hard) "true" jsv_get_param(l_hard) "mem=1G,mem2=200M" jsv_sub_is_param(l_hard,mem) "true" jsv_sub_get_param(l_hard,mem) "1G" jsv_sub_get_param(l_hard,mem3) "false" jsv_sub_get_param(l_hard,mem3) "" Most likely the second to last line should read jsv_sub_is_param instead of jsv_sub_get_param. |
|||
#1296 | wontfix | ensure compatibility with 6.2u5 to pass "standards" test | admin | |
Description |
To be able to use the name "Grid Engine", we have to pass the "standards" test in EXHIBIT B of the SISSL by running the testsuite using a (non-existent) "reference" version (see sge/LICENCES/SISSL, sge/LICENCES/standards, and sge/doc/compatibility_test.html). Although the current version should be compatible with the last of the Sun code, it may not be compatible on the wire with any released version, e.g. because of [3532] and other possible GDI changes in 6.2u6. It would be difficult to argue against compatibility with 6.2u5, as counting, so we need to get the tests running and, if necessary, introduce a compatibility mode that hides changes like [3532] and allows us to pass the test, probably dependent on an environment variable. I don't know how sustainable that will be long-term. |
|||
#1581 | fixed | fails to build from source (#endif without #if) | Afif Elghraoui <afif@…> | afif |
Description |
I attempted to backport your commit to drop use of union wait[1] to the Debian 8.1.9 package, but it causes the following build error (line numbers are relative to the 8.1.9 version): ../clients/qevent/qevent.c: In function ‘qevent_start_trigger_script’: ../clients/qevent/qevent.c:307:2: error: #endif without #if
../clients/qevent/Makefile:83: recipe for target 'qevent.o' failed make[2]: * [qevent.o] Error 1 The attached patch resolves the problem and also streamlines the surrounding ifdefs (otherwise essentially the same condition is being checked three times in a row) |