[needs-packaging] devmem2 needs packaging

Bug #607752 reported by Vincent Stehlé
18
This bug affects 2 people
Affects Status Importance Assigned to Milestone
Ubuntu
Fix Released
Wishlist
Unassigned

Bug Description

There is no Ubuntu package for devmem2.
Could someone please package it?

devmem2 is a tool to read physical memory through /dev/mem
This is useful for low level debugging on embedded targets.
Sources can be found at e.g. http://sources.buildroot.net/devmem2.c

Thanks!

V.

Revision history for this message
Brian Murray (brian-murray) wrote :

*** This is an automated message ***

This bug is tagged needs-packaging which identifies it as a request for a new package in Ubuntu. As a part of the managing needs-packaging bug reports specification, https://wiki.ubuntu.com/QATeam/Specs/NeedsPackagingBugs, all needs-packaging bug reports have Wishlist importance. Subsequently, I'm setting this bug's status to Wishlist.

summary: - devmem2 needs packaging
+ [needs-packaging] devmem2 needs packaging
Changed in ubuntu:
importance: Undecided → Wishlist
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Package created at branch lp:~rsalveti/+junk/devmem2. Could anyone help reviewing it?

For Beagle devmem2 can be used to find the SGX core revision, like:
root@beagle-maverick:~# ./devmem2 0x48004B48 w 0x2
/dev/mem opened.
Memory mapped at address 0x4015c000.
Value at address 0x48004B48 (0x4015cb48): 0x3
Written 0x2; readback 0x2
root@beagle-maverick:~# ./devmem2 0x48004B10 w 0x1
/dev/mem opened.
Memory mapped at address 0x40183000.
Value at address 0x48004B10 (0x40183b10): 0x0
Written 0x1; readback 0x1
root@beagle-maverick:~# ./devmem2 0x48004B00 w 0x2
/dev/mem opened.
Memory mapped at address 0x40160000.
Value at address 0x48004B00 (0x40160b00): 0x0
Written 0x2; readback 0x2
root@beagle-maverick:~# ./devmem2 0x50000014 > sgxrevision.txt
root@beagle-maverick:~# cat sgxrevision.txt
/dev/mem opened.
Memory mapped at address 0x4002a000.
Value at address 0x50000014 (0x4002a014): 0x10201

Depending on the value read out in the last step, OMAPES is determined for the build.

If value == 0x10205, SGX OMAPES=5.x
If value == 0x10201, SGX OMAPES=3.x
If value == 0x10003, SGX OMAPES=2.x

tags: added: arm
Revision history for this message
Michael Opdenacker (michael-opdenacker) wrote :

Hi Ricardo,

I also made my own package for devmem2, but didn't push it any further because my ITP got rejected by Debian.

This experimental package is described on http://omappedia.org/wiki/Building_Software_With_CMake and the package sources can be found on http://free-electrons.com/pub/ubuntu/cmake/devmem2-1.0.0.tar.bz2

Feel free to reuse any of my stuff (the package passed Lintian checks), like the man page...

Cheers,

Michael.

Revision history for this message
Michael Opdenacker (michael-opdenacker) wrote :

I should have added that Debian developers found devmem2 not worthy of packaging, being a quick hack, and not being very portable.

A better solution would be to package BusyBox's devmem, which is an improved implementation.

Would it be possible to add devmem support to the BusyBox package? Should we work with Debian to get this change?

Cheers,

Michael.

Revision history for this message
Michael Opdenacker (michael-opdenacker) wrote :

Oops, I forget to add a link to the Debian bug:http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=595805
Apologies for the excess traffic.

Michael.

Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Maybe getting the improvements from BusyBox's devmem and trying to package it separately would be better. For me it's kind of weird to depend on busybox for it.

Revision history for this message
Michael Opdenacker (michael-opdenacker) wrote :

The advantage of using BusyBox is that devmem would come for free. However, people would have to run "busybox devmem" which is definitely weird (as there are no links to bin/busybox in today's package). So, I agree with you.

Oliver Grawert (ogra)
tags: added: ffe
Revision history for this message
Scott Kitterman (kitterman) wrote :

Why does this need to be pushed for Maverick? Is there any hard requirement that isn't met by waiting for Natty and then doing a backport after it's in?

Changed in ubuntu:
status: New → Incomplete
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Basically because opengles-sgx-omap3 needs it. This package is useful in order to identify the correct SGX hardware.

Currently opengles-sgx-omap3 is deploying 3 different versions of the same libraries, but targeting different hardwares. devmem2 helps finding what is the one that should be used.

Changed in ubuntu:
status: Incomplete → New
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Bug about the PowerVR SGX drivers: bug 652347

Revision history for this message
Allison Randal (allison) wrote :

devmem2 is packaged in universe for maverick and natty, is any further action needed here?

Changed in ubuntu:
status: New → Incomplete
Changed in ubuntu:
status: Incomplete → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.