On May 18, 2014 8:33:15 AM EDT, Francesco Toscan <f.tos...@hotmail.it> wrote: >On Sun, May 18, 2014 at 01:08:00PM +0200, Stefan Sperling wrote: >> On Sun, May 18, 2014 at 12:52:53PM +0200, Francesco Toscan wrote: >> > Hi misc@, >> > >> > Network with bwi has become *slow*, slow as unusable. Every kind of >> > traffic is somehow slowed. Transmissions work but they take forever >to >> > complete. >> >> The CAVEATS section of the man page seems to describe your issue. >> Perhaps it's a hardware problem? Does it work again if you downgrade >> to 5.5-release? > >Stefan, > >thanks for your reply. > >I thought my device was one among those fortunate working chips :)) >I'm looking at bwi.c 's CVS record: last commit was 8 weeks ago, I'm >sure performances were suboptimal before that. >Preceeding commit was 5 months ago and I'm sure in February 26th the >driver performed better. >This snapshot from May 12th performs better than three/four weeks ago, >but the driver was not modified: can't give numbers but three/four >weeks >ago network was stuck, now it's just very slow. >Unfortunately between last February and April I was forced to use this >laptop as a rescue FreeBSD 9 system (btw, bwi worked fine there) so I >have an usage blackout of two months. >I'll boot 5.5-release and report what's happening. >-- >f.
Not as useful a report as this could be, but I can confirm this regression on a Dell D630 - I initially thought it was just the crappy conference WiFi at BSDCan, but my windows laptop and my phone had no difficulty when operating side-by-side in the same band. In my case, 5.3-RELEASE worked fine but 5.5-RELEASE behaved as the OP describes: unusably slow. (By unusably, I mean that I couldn't reliably use pkg_add with a remote PKG_PATH to upgrade packages - it kept dying on partial packages downloads.) Dmesg and diag might follow later today if I can, I'm traveling. -Adam -- Sent from my Android device with K-9 Mail. Please excuse my brevity.
