ports/181584: x11/nvidia-driver NVIDIA-FreeBSD-x86_64-319.32 fails to build due to uninitialized data warning and -Werror flag
Nathanael Hoyle
nhoyle at hoyletech.com
Tue Aug 27 14:20:00 UTC 2013
>Number: 181584
>Category: ports
>Synopsis: x11/nvidia-driver NVIDIA-FreeBSD-x86_64-319.32 fails to build due to uninitialized data warning and -Werror flag
>Confidential: no
>Severity: non-critical
>Priority: low
>Responsible: freebsd-ports-bugs
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: current-users
>Arrival-Date: Tue Aug 27 14:20:00 UTC 2013
>Closed-Date:
>Last-Modified:
>Originator: Nathanael Hoyle
>Release: 9.1-RELEASE
>Organization:
>Environment:
FreeBSD tsuba 9.1-RELEASE FreeBSD 9.1-RELEASE #0 r243825: Tue Dec 4 09:23:10 UTC 2012 root at farrell.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC amd64
>Description:
Build fails with:
nvidia_acpi.c: In function 'nv_acpi_dsm_method':
nvidia_acpi.c:166: warning: 'data_size' may be used uninitialized in this function
*** [nvidia_acpi.o] Error code 1
note that this is a warning, but -Werror compiler flag is forced on (not present in make.conf).
Subsequent message recommends setting MAKE_JOBS_UNSAFE=yes, but this has no effect.
>How-To-Repeat:
Attempt to build x11/nvidia-driver from ports.
>Fix:
The first usage of the data_size variable appears to be on line 203, and it is used as an "out" argument to nvidia_acpi_extract_object. That function initializes it to 0 at the beginning (line 97), so the usage is not actually unsafe. Nevertheless, initializing it to 0 at the point of declaration should have no change in behavior and suppress this warning.
--- nvidia_acpi.c.orig 2013-08-27 10:11:43.149516389 -0400
+++ nvidia_acpi.c 2013-08-27 10:12:02.588515738 -0400
@@ -163,7 +163,7 @@
union acpi_object *dsm;
union acpi_object dsm_input[4];
struct acpi_object_list input = { 4, dsm_input };
- uint32_t data_size;
+ uint32_t data_size = 0;
dev = device_get_parent(nv->handle);
>Release-Note:
>Audit-Trail:
>Unformatted:
More information about the freebsd-ports-bugs
mailing list