Returned original module name. #34
Conversation
This one is always selected. Signed-off-by: Larry Finger <[email protected]>
This variable is always selected. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
The standard kernel features can be used to implement bridges. Signed-off-by: Larry Finger <[email protected]>
…VE_CHECK Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
…dard Signed-off-by: Larry Finger <[email protected]>
Since converting to CFG80211, warnings of the following type have been issued: [88526.660804] WARNING: CPU: 3 PID: 9705 at net/wireless/sme.c:848 __cfg80211_disconnected+0x3b8/0x3e0 [cfg80211]() Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
All of this functionality can be handled by cfg80211. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Thhis hack is replaced by traditional operations on the skb. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
The original driver is capable of about 80 Mbps when running netperf TCP tests. Commit d88f1fa introduced an error that cut the throughput to about 30 Mbps. These changes improve the results to RX of 46 and TX of 60 Mbps. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
This patch fixes the remaining bug that was limiting the 5G rate. Both RX and TX get in excess of 80 Mbps using netperf. In fact, the actual rate could be higher as my netperf server is connected to the AP with a 100 Mbps wired connection, and netperf only gets about 90 Mbps over the wire. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
For some reason, an extra parameter was added to the call to cfg80211_rx_mgmt() in kernel 3.17, and then removed in 3.18. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
There are two warnings left. Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
This reverts commit 242ffe1.
A warning that appears with kernel 4.1 is also fixed. Signed-off-by: Larry Finger <[email protected]>
Some dkms.conf cleanup
|
This change is OK, but I will not merge it as it violates my naming convention. Names rtlxxxxyy are reserved for drivers in the regular wireless tree. Names rxxxxyy are for drivers in the staging tree, and names xxxyy are for drivers that are not merged into any kernel tree. Why did you need to change the name? |
|
When installing with dkms it is not good to change original names or blacklist original modules. |
|
Correct, there is no rtl8192du in the kernel, and there is never likely to be. That would require a complete rewrite of the driver to use mac80211 instead of the current MAC stack written by Realtek. If the current driver were to be added to the kernel, it would be called r8192du. |
|
OK. I just pushed that driver to Ubuntu ppa. I will think how to get it smooth with namings. It is good to have all updates in your ppa, so I could just pull and repack them. Let's leave it as it is for a while then. |
lwfinger
force-pushed
the
master
branch
2 times, most recently
from
e4f3e9a
to
f20ede8
Compare
Some dkms.conf cleanup