Messages in this thread | | | Date | Fri, 22 Jan 1999 11:59:08 -0500 | From | Randy Gobbel <> | Subject | 2.2.0-"final": aic7xxx still doesn't work with Promise Ultra33 (etc.) |
| |
Not surprising since Doug Ledford is away, but nothing has changed in recent versions of the aic7xxx driver, and it still has serious problems. Trying to use an Adaptec 2940UW SCSI card together with a Promise DC20246 Ultra33 IDE adapter results in either a hard freeze or a bunch of PCI Master Abort errors, depending on which driver is started first. It is unclear if the root of the problem is really in aic7xxx, the IDE code, or somewhere down in the BIOS code. This setup worked fine with the Promise Ultra33 patch in 2.0.35 (and I believe DC20246 support was incorporated by default in later 2.0 kernels).
In addition, there are problems with several types of SCSI tape drives, which hang up and cannot be reset by anything short of a reboot. The problem is said to be in st.c, but that's all I know. The point is that this is another facility that worked fine in earlier kernels, and is now seriously broken.
Finally, there are still some oddities in the xconfig menus, like having the QIC-02 tape driver and watchdog timer support items show up under "Mice".
I don't know about Linus, but I would be embarrassed to release a kernel with that many formerly-working things broken in it. If I were a Microsoft publicist, I would be rubbing my hands in anticipation of the excellent FUD fodder in the new Linux kernel. New features and new devices are one thing, but breaking stuff that used to work--particularly if it prevents systems from booting, like the aic7xxx-DC20246 conflict does--should be an absolute no-no.
How about having a *real* code freeze, as in: the only fixes allowed are those that deal with disabling problems in subsystems that worked properly in 2.0.xx? Things like swapper optimizations and fixes to platforms and devices that were not supported previously don't qualify.
-Randy -- http://www.cnbc.cmu.edu/~gobbel/ PGP fingerprint: 32 8A E8 24 A1 46 26 BC F9 9D 0E B6 81 A9 02 0C
NOTICE: I DO NOT ACCEPT UNSOLICITED COMMERCIAL EMAIL MESSAGES OF ANY KIND.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |