[Bug 100011] New: src/intel_device.c reverses return value meaning of xorg server xf86LoadKernelModule()
bugzilla-daemon at freedesktop.org
bugzilla-daemon at freedesktop.org
Tue Feb 28 20:29:24 UTC 2017
https://bugs.freedesktop.org/show_bug.cgi?id=100011
Bug ID: 100011
Summary: src/intel_device.c reverses return value meaning of
xorg server xf86LoadKernelModule()
Product: xorg
Version: git
Hardware: Other
OS: FreeBSD
Status: NEW
Severity: normal
Priority: medium
Component: Driver/intel
Assignee: chris at chris-wilson.co.uk
Reporter: davshao at gmail.com
QA Contact: intel-gfx-bugs at lists.freedesktop.org
The function load_i915_kernel_module() in file src/intel_device.c
of xf86-video-intel reverses the meaning of the return value of
xorg server's xf86LoadKernelModule(). xf86LoadKernelModule()
actually returns 0 if the loading fails, non-zero if success.
For example, from xorg server 1.19.1's source code,
file hw/xfree86/os-support/linux/lnx_kmod.c:
* Input:
* modName - name of the kernel module (Ex: "tdfx")
* Return:
* 0 for failure, 1 for success
*/
int
xf86LoadKernelModule(const char *modName)
if (WIFEXITED(status) && WEXITSTATUS(status) == 0) {
return 1; /* success! */
}
else {
return 0;
}
And from file hw/xfree86/os-support/bsd/bsd_kmod.c
* Return:
* 0 for failure, 1 for success
*/
int
xf86LoadKernelModule(const char *modName)
{
if (kldload(modName) != -1)
return 1;
else
return 0;
}
On FreeBSD there is an additional complication that if
"i915kms" is added to kernel_module_names, the inverted use
of the return value causes both i915kms.ko and i915.ko to
be loaded.
Therefore I believe a patch similar to the following needs to be made
to src/intel_device.c:
@@ -233,9 +242,8 @@ static int load_i915_kernel_module(void)
const char **kn;
for (kn = kernel_module_names; *kn; kn++)
- if (xf86LoadKernelModule(*kn) == 0)
+ if (xf86LoadKernelModule(*kn))
return 0;
-
return -1;
}
--
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20170228/7d9dcd6f/attachment.html>
More information about the intel-gfx-bugs
mailing list