Hey Marc-André, Thanks for the review.
On 03/16/2015 11:22 AM, Marc-André Lureau wrote: > Hi Jeremy > > I have checked no regression for spice, but not the new pcsc backend. > Do you have a spice-gtk branch or are you using vscclient only? I'm using spice-gtk with this patch: http://lists.freedesktop.org/archives/spice-devel/2015-January/018674.html and then invoking it with --spice-smartcard[=passthru] > > I would feel better note polluting more the exported symbols from > libcacard. For ex, vcard_init() and cac_card_init() are already quite > useless, since they are called from vcard_emul_init() Would you be > able to init PCSC there too, when it's selected? (there are too many > init or exported functions already) > > Why did you export vcard_response_set_status_bytes()? > It seemed as though the convention was that all non static functions went into the .syms file; I may have misunderstood. capcsc_init certainly does not need to be exported; it is only intended to be used from vcard_emul_init(). I thought I was simply following existing convention. For vcard_response_set_status_bytes, I needed that function in capcsc.c. Again, I followed what I saw as the convention. However, that one is perhaps arguably more useful to export; certainly if you're going to export the other vcard_response_xxx functions, this would be a useful one to add to the mix. Cheers, Jeremy
