small modification,, added -B option t glxinfo to get rid of unneeded output.

Also found a case with intel gpu where glxinfo -display :0 hangs endlessly when out of X.

I assume this is an Intel or glx bug, so I mentioned this in the man page for -! 40 option.
This commit is contained in:
Harald Hope 2017-06-08 20:14:16 -07:00
parent 8d651eb345
commit be5714c8b3
2 changed files with 5 additions and 6 deletions

8
inxi
View file

@ -2078,7 +2078,8 @@ debug_data_collector()
touch $debug_data_dir/xprop-absent
fi
if type -p glxinfo &>/dev/null; then
glxinfo &> $debug_data_dir/glxinfo.txt
glxinfo &> $debug_data_dir/glxinfo-full.txt
glxinfo -B &> $debug_data_dir/glxinfo-B.txt
else
touch $debug_data_dir/glxinfo-absent
fi
@ -6407,7 +6408,7 @@ get_graphics_glx_data()
# if [[ $B_SHOW_DISPLAY_DATA == 'true' && $B_ROOT != 'true' ]];then
if [[ $B_SHOW_DISPLAY_DATA == 'true' ]];then
IFS=$'\n'
A_GLX_DATA=( $( eval glxinfo $DISPLAY_OPT 2>/dev/null | gawk -F ': ' '
A_GLX_DATA=( $( eval glxinfo -B $DISPLAY_OPT 2>/dev/null | gawk -F ': ' '
# the real question here though is why this function is even here, seems
# just to be a complicated way to pack/print a variable, but maybe the
# original idea was to handle > 1 cases of detections I guess
@ -6467,9 +6468,6 @@ get_graphics_glx_data()
/direct rendering/ {
$2 && c[$2]
}
/GLX Visuals/ {
exit
}
END {
dr = join( c, ", " )
oglr = join( a, ", " )

3
inxi.1
View file

@ -535,7 +535,8 @@ Force use of \fBdmidecode\fR. This will override \fB/sys\fR data in some lines,
.B \-! 40
Will try to get display data out of X. Default gets display info from display \fB:0\fR. If you use this format: \fB\-40:1\fR
it would get it from display \fB1\fR instead, or any display you specify as long as there is no space
between \fB\-40\fR and the \fB:[display id]\fR."
between \fB\-40\fR and the \fB:[display id]\fR. Note that in some cases, this will cause inxi to hang endlessly, with Intel
graphics this has been seen, so it won't work for you in that case.
.SH DEBUGGING OPTIONS
.TP