Comment 4 for bug 385703

My point exactly. Thanks for helping me clarify this bug even more. Since no one i paying attention, I was prepared to make a patch myself, thou my idea was to modifying the way "nv.ids" is generated in "01_gen_pci_ids.diff".

A the moment the script looks like this:

awk '/{ 0x.*/ || /case 0x.*/ {print $$2}' ${srcdir}/nv_driver.c | sed -e s/0x// -e s/,// -e s/:// -e s/^0/10DE0/ | sort -u > nv.ids

This line doesn't even work and "nv.ids" must be cleaned afterwards. "/case 0x.*/" is useless too. My idea was to change it to something like this:

awk '/{ 0x.*/ {print $$2}' nv_driver.c | sed -e 's/ { //' -e 's/0x//' -e 's/,.*//' -e 's/:.*//' > nv_tm
p.ids; for i in 0 1 2 3 4 5 6 7 8 9 A B C D E F; do echo "10DE00F$i"; done >> nv_tmp.ids; for i in 0 1 2 3 4; do echo "10DE02E$i"; done >> nv_tmp.ids; sort
-u nv_tmp.ids > nv.ids

There is probably more elegant way but I'm not realy good at bash :).

In, the end it seems that it doesn't even matter any more since *.ids are going to be dropped.