On Tue, 2 May 2006, Sancho Dauskardt wrote: > Hi all, > > when heavily pollig usb-storage devices with a TEST_UNIT_READY > command and then removing the device occasionally causes an Oops near khubd. > Tried on Kernel 2.6.11.1 and 2.6.16.11. > > To reproduce, run following on a multi-lun card-reader: > ----- snip -------- > while [ 1 ]; do > sg_turs /dev/sg0; > sg_turs /dev/sg1; > sg_turs /dev/sg2; > sg_turs /dev/sg3; > done > ----- snip -------- > and then remove/insert the device a few times. > > 'Best' result on a single-cpu machine without udev, etc > (Hyperthreading SMP also oopses). > All my boxes will hang after 3-15 remove/insert cycles (SIS and VIA ehci). > > So far only USB2.0 devices where tested (could it matter ?). > > Any pointers ?
This kind of stress testing is a great thing to do. >From your logs, the problems you encounter are part of the SCSI stack -- the sg driver to be specific. Try running on the most recent 2.6.17-rc kernel, and post your findings to [EMAIL PROTECTED] Maybe CC: linux-usb-devel also. Alan Stern ------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
