lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
Open Source and information security mailing list archives
| ||
|
Message-ID: <6b725e37-4223-cef8-5430-aa05ddafb551@gmail.com> Date: Mon, 31 Dec 2018 22:48:49 +1300 From: Michael Schmitz <schmitzmic@...il.com> To: David Miller <davem@...emloft.net> Cc: alex@...ik.de, netdev@...r.kernel.org, linux-m68k@...r.kernel.org, rolf.anders@...sik.uni-augsburg.de Subject: Re: [PATCH net-next] Amiga PCMCIA 100 MBit card support Hi Dave, Am 29.12.2018 um 09:43 schrieb David Miller: > From: Michael Schmitz <schmitzmic@...il.com> > Date: Sat, 22 Dec 2018 10:30:58 +1300 > >> Am 21.12.2018 um 13:10 schrieb David Miller: >>> And in particular this huge complicated Kconfig construct is not >>> maintainable at all. >> >> We can trim this down a bit (for reasons I've outlined before): >> >> +if APNE >> +config APNE100MBIT >> + bool "PCMCIA NE2000 100MBit support" >> + default n >> + depends on NE2000=n && PCMCIA_AXNET=n >> + depends on PCMCIA_PCNET=n && STNIC=n && ULTRA=n && WD80x3=n >> >> In all honesty, I doubt anyone could ever use the stnic, ultra and wd >> drivers on m68k (Geert?) so these could be omitted as well. Not sure >> pcnet_cs or axnet_cs are possible to use on the Amiga PCMCIA slot, so >> all that remains in practice is the ne driver (which is used on >> Atari). STNIC can indeed be dropped from the list of exceptions, but I forgot about ISA support on Q40. >> >> Still too ugly? > > I'm sorry, maybe I didn't express my concern clearly. I understood your concerns - my hope was that the exception list might be possible to trim right down to a single driver. Alas, that's not the case. > Any sizable list of exceptions in a Kconfig conditional has long term > maintainence costs. If, for some reason, a new variant of support > for this chip arrives there will be a new Kconfig setting that would > need to be added here to your expressions. > > That is insanely error prone, and I can guarantee it will get missed. > > Having so many front end drivers for a chipset should be done in a > more modular manner, so that in fact they could all coexist if > necessary. Not sure I follow your meaning ... are you suggesting to allow the back end driver to use run-time selectable IO accessors and other generic helper functions, so the drivers currenlty including lib8390.c directly could use the core code from the 8390 module instead? Cheers, Michael
Powered by blists - more mailing lists