On Tue, 25 Apr 2006, Christian Haul wrote: > Hi all, > > I've recently noticed a problem with extremely high system load on my > debian sid system. I first tracked it down to the usb_storage kernel > module ie removing the module fixes the problem. Now, I get the > impression that hald triggers the > problem by continuesly polling the card reader device. > I've compiled the usb module with debug on and the following shows up in > /var/log/debug: > > Apr 22 23:16:51 polo kernel: usb-storage: queuecommand called > Apr 22 23:16:51 polo kernel: usb-storage: *** thread awakened. > Apr 22 23:16:51 polo kernel: usb-storage: Command TEST_UNIT_READY (6 bytes) > Apr 22 23:16:51 polo kernel: usb-storage: 00 00 00 00 00 00 > Apr 22 23:16:51 polo kernel: usb-storage: Bulk Command S 0x43425355 T > 0x128c3 L 0 F 0 Trg 0 LUN 0 CL 6 > Apr 22 23:16:51 polo kernel: usb-storage: usb_stor_bulk_transfer_buf: > xfer 31 bytes > Apr 22 23:16:51 polo kernel: usb-storage: Status code 0; transferred 31/31 > Apr 22 23:16:51 polo kernel: usb-storage: -- transfer complete > Apr 22 23:16:51 polo kernel: usb-storage: Bulk command transfer result=0 > Apr 22 23:16:51 polo kernel: usb-storage: Attempting to get CSW... > Apr 22 23:16:51 polo kernel: usb-storage: usb_stor_bulk_transfer_buf: > xfer 13 bytes > Apr 22 23:16:51 polo kernel: usb-storage: Status code 0; transferred 13/13 > Apr 22 23:16:51 polo kernel: usb-storage: -- transfer complete > Apr 22 23:16:51 polo kernel: usb-storage: Bulk status result = 0 > Apr 22 23:16:51 polo kernel: usb-storage: Bulk Status S 0x53425355 T > 0x128c3 R 0 Stat 0x1 > Apr 22 23:16:51 polo kernel: usb-storage: -- transport indicates command > failure > Apr 22 23:16:51 polo kernel: usb-storage: Issuing auto-REQUEST_SENSE > Apr 22 23:16:51 polo kernel: usb-storage: Bulk Command S 0x43425355 T > 0x128c4 L 18 F 128 Trg 0 LUN 0 CL 6 > Apr 22 23:16:51 polo kernel: usb-storage: usb_stor_bulk_transfer_buf: > xfer 31 bytes > Apr 22 23:16:51 polo kernel: usb-storage: Status code 0; transferred 31/31 > Apr 22 23:16:51 polo kernel: usb-storage: -- transfer complete > Apr 22 23:16:51 polo kernel: usb-storage: Bulk command transfer result=0 > Apr 22 23:16:51 polo kernel: usb-storage: usb_stor_bulk_transfer_buf: > xfer 18 bytes > Apr 22 23:16:51 polo kernel: usb-storage: Status code 0; transferred 18/18 > Apr 22 23:16:51 polo kernel: usb-storage: -- transfer complete > Apr 22 23:16:51 polo kernel: usb-storage: Bulk data transfer result 0x0 > Apr 22 23:16:51 polo kernel: usb-storage: Attempting to get CSW... > Apr 22 23:16:51 polo kernel: usb-storage: usb_stor_bulk_transfer_buf: > xfer 13 bytes > Apr 22 23:16:51 polo kernel: usb-storage: Status code 0; transferred 13/13 > Apr 22 23:16:51 polo kernel: usb-storage: -- transfer complete > Apr 22 23:16:51 polo kernel: usb-storage: Bulk status result = 0 > Apr 22 23:16:51 polo kernel: usb-storage: Bulk Status S 0x53425355 T > 0x128c4 R 0 Stat 0x0 > Apr 22 23:16:51 polo kernel: usb-storage: -- Result from auto-sense is 0 > Apr 22 23:16:51 polo kernel: usb-storage: -- code: 0x70, key: 0x2, ASC: > 0x3a, ASCQ: 0x0 > Apr 22 23:16:51 polo kernel: usb-storage: Not Ready: Medium not present > Apr 22 23:16:51 polo kernel: usb-storage: scsi cmd done, result=0x2 > Apr 22 23:16:51 polo kernel: usb-storage: *** thread sleeping. > > This block repeats for every 2 seconds..... and top shows around CPU 40% hi
I have trouble believing that a few milliseconds of activity occurring once every two seconds is enough to push your CPU usage to 40%. > This is what I got from [EMAIL PROTECTED] : > > > We need to poll otherwise we can't do media detection. The kernel driver > > needs to be fixed, suggest to write mail to one of the kernel lists > > detailing the problem. They're right about the polling, but they're wrong about the kernel driver needing to be fixed. In fact nothing needs to be fixed as far as I can see, except possibly the statistics you get from top. Have you tried experimenting with a real load present, such as running a benchmark program? Try to find out how much hald and usb-storage really do slow down performance. On the other hand, if you _really_ object to the extra load, you can simply turn off hald. Or rmmod usb-storage. 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-users@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-users