Groups > linux.debian.bugs.dist > #883265

Bug#891036: Proposal - how to make usbip more resistant to misconfiguration

From	"wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl>
Newsgroups	linux.debian.bugs.dist, linux.kernel, linux.debian.kernel
Subject	Bug#891036: Proposal - how to make usbip more resistant to misconfiguration
Date	2018-02-23 02:10 +0100
Message-ID	<vma25-4Ff-5@gated-at.bofh.it> (permalink)
References	<vlZgm-5tv-5@gated-at.bofh.it> <vm7nA-2DF-23@gated-at.bofh.it> <vm7QB-2QM-3@gated-at.bofh.it> <vlH06-1O8-3@gated-at.bofh.it> <vm7QB-2QM-3@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Cross-posted to 3 groups.

Show all headers | View raw

Hi,

I have found that the
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=891036 is not the only
one related to misconfiguration of the usbip driver and tools.
A few weeks ago the similar bug
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=878866 was discovered.
It seems that the configuration of usbip is fragile, and it is easy to
make a mistake leading to the setup that doesn't work correctly.
I suppose, that additional mechanisms should be added to prevent such
problems:

 1. It should be checked DURING THE COMPILATION, that the number of
    ports in a single virtual hub does not exceed 128.
 2. The MAXNPORT constant in tools/usb/usbip/libsrc/vhci_driver.h should
    be defined as:
    #define MAXNPORT ( VHCI_PORTS * VHCI_NR_HCS )

I have not tested the above modifications. If I find some time to verify
them, and if they are correct, I'll submit the official patches.

With best regards,
Wojtek

-- 
Wojciech M Zabolotny, PhD
Institute of Electronic Systems
Faculty of Electronics and Information Technology
Warsaw University of Technology

Back to linux.debian.bugs.dist | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

Bug#891036: libusbip: error: port number exceeds 128 wzab <wzab@ise.pw.edu.pl> - 2018-02-21 19:10 +0100
  Bug#891036: More details - the problem is associated with the vhci-hcd driver. wzab <wzab@ise.pw.edu.pl> - 2018-02-21 20:00 +0100
  Bug#891036: Problem occurs in 4.13 and 4.14, but not in 4.12 "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-21 20:30 +0100
  Bug#891036: Problem still persists in kernel 4.15 "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-21 21:10 +0100
  Bug#891036: Even after updating "usbip" to usbip_2.0+4.15.4-1_amd64.deb problem remains "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-21 21:20 +0100
  Bug#891036: The problem is related to the number of ports created in vhci_hcd.0 Wojtek Zabolotny <W.Zabolotny@elka.pw.edu.pl> - 2018-02-22 14:40 +0100
    Bug#891036: The problem is related to the number of ports created in vhci_hcd.0 "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-22 23:20 +0100
      Bug#891036: The problem is related to the number of ports created in vhci_hcd.0 "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-22 23:50 +0100
        Bug#891036: Workaround found, the problem is with Debian configuration of kernel "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-23 01:10 +0100
        Bug#891036: Proposal - how to make usbip more resistant to misconfiguration "wzabolot@elektron.elka.pw.edu.pl" <wzabolot@elektron.elka.pw.edu.pl> - 2018-02-23 02:10 +0100

csiph-web