Messages in this thread | | | Date | Fri, 22 Jan 1999 18:03:09 -0500 (EST) | From | "Adam D. Bradley" <> | Subject | Re: 2.2.0-final |
| |
On Wed, 20 Jan 1999, Linus Torvalds wrote:
> In short, before you post a bug-report about 2.2.0-final, I'd like you to > have the following simple guidelines: > > "Is this something Linus would be embarrassed enough about that he would > wear a brown paper bag over his head for a month?" > > and > > "Is this something that normal people would ever really care deeply > about?"
Don't know if this counts... depends on how many people out there are using 3c905B Cyclone cards...
The 3c509 update that went in in 2.2.0-pre6 is giving me trouble. Specifically, it seems to be causing NFS (and any other high-bandwidth app, like f.e. scp) to suck eggs.
Running "strace cat big-file-on-irix6.5-nfs-server > /dev/null" shows bursts of 19 successful 4096-byte reads, between which the process sleeps for 1-8 seconds in a read call. That 19 is _very_ consistent.
Running ifconfig shows the interface reporting errors at a rate of roughly one per second when the network is under load. After a few minutes:
eth0 Link encap:Ethernet HWaddr 00:10:4B:0D:F8:E5 inet addr:128.197.14.46 Bcast:128.197.14.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:6380 errors:193 dropped:0 overruns:0 frame:194 TX packets:1716 errors:0 dropped:0 overruns:0 carrier:0 Collisions:0 Interrupt:17 Base address:0xe400
Compare this with stressing it for <30secs under pre5:
eth0 Link encap:Ethernet HWaddr 00:10:4B:0D:F8:E5 inet addr:128.197.14.46 Bcast:128.197.14.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:16902 errors:0 dropped:0 overruns:0 frame:0 TX packets:5667 errors:0 dropped:0 overruns:0 carrier:2 Collisions:14 Interrupt:17 Base address:0xe400
Driver initialization reports:
3c59x.c:v0.99H 11/17/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html eth0: 3Com 3c905B Cyclone 100baseTx at 0xe400, 00:10:4b:0d:f8:e5, IRQ 17 8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface. MII transceiver found at address 24, status 786d. MII transceiver found at address 0, status 786d. Enabling bus-master transmits and whole-frame receives.
Compared with pre5:
eth0: 3Com 3c905B Cyclone 100baseTx at 0xe400, 00:10:4b:0d:f8:e5, IRQ 17 8K byte-wide RAM 5:3 Rx:Tx split, autoselect/NWay Autonegotiation interface. Enabling bus-master transmits and whole-frame receives. 3c59x.c:v0.99E 5/12/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html
These problems are consistent across pre6-7-8-9.
In case it matters, the system is a dual-PII and I'm on a 100Mb switch.
Sorry this took a few days to report - the problem only became apparent when I started doing high-bandwidth stuff over NFS, and I have been too busy teaching and doing linux propoganda[1] to properly isolate the problem until today.
Adam [1] - Gave a talk for CONE, "Computer Organizations of New England" -- Your lives aren't small, but \\ Adam Davenport Bradley, Grad Student you're living them in a small \\ Boston University Computer Science way. Live openly and expansively! \\ artdodge@cs.bu.edu 353-8921/MCS211 II Cor 6:12-13 (The Message) <>< \\ http://www.netwinder.org/~artdodge
- 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/
| |