I have an old PII nfs media server running 2.6.10-rc3 with an intel
e100. Its interrupt is shared with both controllers of an siimage PCI
ide card:
[dunbar](1) # cat /proc/interrupts
CPU0
0: 16835371 XT-PIC timer
2: 0 XT-PIC cascade
8: 4 XT-PIC rtc
10: 75 XT-PIC uhci_hcd
11: 1685559 XT-PIC ide2, ide3, eth0
15: 0 XT-PIC ide1
NMI: 0
LOC: 0
ERR: 0
With the eepro100 driver, a single tcp stream goes wirespeed in either
direction (11.2MB/s). With e100, the speed fluctuates wildly without
ever exceeding 2.5MB/s or so, and it averages around 1.
I thought it might be due to an interaction of napi and the shared
interrupt, so I tried disabling napi in the driver. Which was
frustrating, because it's no longer optional, despite what Kconfig says.
Jason
Signed-off-by: Jason Lunz <lunz@xxxxxxxxxxxx>
--- linux-2.6.10-rc3/drivers/net/Kconfig.pre 2004-12-11 23:11:49.000000000
-0500
+++ linux-2.6.10-rc3/drivers/net/Kconfig 2004-12-11 23:12:13.000000000
-0500
@@ -1435,23 +1435,6 @@
<file:Documentation/networking/net-modules.txt>. The module
will be called e100.
-config E100_NAPI
- bool "Use Rx Polling (NAPI)"
- depends on E100
- help
- NAPI is a new driver API designed to reduce CPU and interrupt load
- when the driver is receiving lots of packets from the card. It is
- still somewhat experimental and thus not yet enabled by default.
-
- If your estimated Rx load is 10kpps or more, or if the card will be
- deployed on potentially unfriendly networks (e.g. in a firewall),
- then say Y here.
-
- See <file:Documentation/networking/NAPI_HOWTO.txt> for more
- information.
-
- If in doubt, say N.
-
config LNE390
tristate "Mylex EISA LNE390A/B support (EXPERIMENTAL)"
depends on NET_PCI && EISA && EXPERIMENTAL
|