New branch 'private/jmux/oss-fuzz-wip' available with the following commits:
commit 737e248da0b01347174770dd274217bf64c95516
Author: Jan-Marek Glogowski <[email protected]>
Date: Mon Dec 13 13:30:46 2021 +0100
local ossfuzz changes
Change-Id: I5b35e0464fd38f85bf5682c40fa7e1f8ab81ef9e
commit d1e172462c0e89aa3631f3fcb26e8bc3a1073828
Author: Jan-Marek Glogowski <[email protected]>
Date: Mon Dec 13 12:10:43 2021 +0100
gbuild: forward / set linktargetmakefilename
Unfortunatly RepositoryExternal.mk ais generally not forwarding
this info. It's also not really documented.
Change-Id: I7920f7442521ab06f18c16397a13615e4517d053
commit 00b4ca87e42add14b2eaaa5e34df84382e0a44de
Author: Jan-Marek Glogowski <[email protected]>
Date: Mon Dec 13 12:05:33 2021 +0100
Respect verbose setting for module-deps.pl calls
Can't use $(gb_VERBOSE), as this runs pre-gbuild.
Change-Id: I7343ede538a246d54f3ce9fcaf104dfbdea08653
commit f593ee5e28fae537d7f919b27701f278475a93e1
Author: Jan-Marek Glogowski <[email protected]>
Date: Sun Dec 12 22:13:37 2021 +0100
Fail module-deps.pl on more errors
* Fail on the error code returned from make
* Collect dependency errors in the tree and print them
The "perldoc -f open" has an example in the "Opening a filehandle
into a command" section, which uses waitpid in the parent. Since
we parse the pipe
I'm not sure the waitpid actually works, as the process has
already ended at this point and the STDOUT pipe from make is
closed. So waitpid actually returns -1 here, as the process
is already gone. From "perldoc -f open" it is not clear, what
exactly happens: if the child calls execve, then the exit
wouldn't make any sense. The STDOUT pipe is just
open in the parent process. The whole setup
Change-Id: I2e77aa33a0eaa5d6d3e06904bb0af53f4b66ef91
commit 21668056cc9935ac66a54e00b8cdc96924037cb0
Author: Jan-Marek Glogowski <[email protected]>
Date: Sun Dec 12 22:03:49 2021 +0100
oss-fuzz: convert to static cross-build
This way we can get rid of the pre-generated locale data.
Change-Id: I3aed9ec62da507254b42e4e7470bae1097b4bc8c
commit 71bdca7e6000fcbd93d1bc60643fe5ee1a1570b0
Author: Jan-Marek Glogowski <[email protected]>
Date: Mon Dec 6 17:41:30 2021 +0000
lockfile: don't serialize locking of lockfile
Change-Id: Ie4498e12b3d0f676ed0c9abf4b3bb4899d6a1c03
commit 176d1783a83120a99adbdf6f9a2366c728e23a24
Author: Jan-Marek Glogowski <[email protected]>
Date: Fri May 21 15:41:15 2021 +0200
gbuild: build static LO / link static executables
See the (large) comment in solenv/gbuild/static.mk trying to
explain, why this implementation was chosen (spoiler: seems
there is no other way) and what is actually implemented. Yes,
I also think it's borderline maintainable (like gbuild in
general; complexity clashing with make "restrictions").
I which I had put that much time into a Meson build, or just
had expanded the bin/lo-all-static-libs "concept"...
Change-Id: Iafc95752fae9e88095f54a21f1e30a4f080815e2
commit 11dca0c4f611b7052ee9648bc5c73ea3d6782180
Author: Jan-Marek Glogowski <[email protected]>
Date: Mon Dec 13 11:54:32 2021 +0100
Refactor some buildsytem stuff
Change-Id: Idfca786ecc7251e08525bd5b45936143727c43d3
