Hi Dave
I've managed to reproduce a relocation error when loading a module built
for Thumb. The runtime error message is:
section 4 reloc 247 sym '': relocation 30 out of range (0xbf9710c6 ->
0xbf96c000)
Relocation 30 is R_ARM_THM_JUMP24, and looking at apply_relocate() in
arch/arm/kernel/module.c it looks like the error is caused by 'offset'
being even here ...
/* only Thumb addresses allowed (no interworking) */
if (!(offset & 1) ||
offset <= (s32)0xff000000 ||
offset >= (s32)0x01000000) {
pr_err("%s: section %u reloc %u sym '%s': relocation %u out of range (%#lx -> %#x)\n",
module->name, relindex, i, symname,
ELF32_R_TYPE(rel->r_info), loc,
sym->st_value);
return -ENOEXEC;
}
I seem to remember you posting some similar issue to a mailing list
recently, or is my memory faulty?
I have some labels declared in inline assembler, and after adding
.type name_of_label, %function
the relocation error goes away, presumably because the label is now
thumbified by having bit 0 set.
--
Tixy
== Shawn Guo (shawnguo) ==
=== Device Tree ===
* Wolfram gave partial comments on sdhci-pltfm&OF-consolidation,
and still wait for more to start v3. Mostly, the patch #5 which
consolidates imx and mpc esdhc support will get dropped.
=== LTP Blueprint ===
* Drafted work items
https://blueprints.launchpad.net/linux-linaro/+spec/linaro-kernel-o-ras-fix…
=== Misc ===
* Tested RMK's MMIO-clocksources-consolidation patch on mxs
* Tested Wolfram's rtc-stmp37xx patch on mxs
* Posted the first try of moving mxs gpio driver into drivers/gpio
=== Plan ===
* Try to move mxc gpio driver into drivers/gpio
--
Regards,
Shawn
== Paul E. McKenney <paulmck> ==
=== Highlights ==
* Attended Budapest summit, led maintainership and Android meetings.
* Blogged maintainership meetings and published LWN article (http://lwn.net/Articles/443510/).
* RCU priority-boosting commits accepted into mainline for 2.6.40.
* Cheng Xu found a fix for the related scheduler bug, which is now in mainline.
* Frederic Weisbecker found a fix for a bug in lock self-test that was resulting in RCU grace-period hangs, which I queued up in my -rcu tree.
* Compiler-sensitive slowdowns located by Yinghai Lu, currently tracking this down.
== Miscellaneous ==
* Handled a license request.
* Discussed the possibility of heavy-duty RCU correctness tools with researchers at Queen Mary University in London.
* Discussed the possibility of formalizing Linux-kernel memory-barrier semantics with researchers at Cambridge University in Cambridge.
* Queued a modification to RCU's fast-path in response to a code-generation complaint from Linus Torvalds.
== Per Forlin <perfor> ==
=== Highlights ===
* Updating blueprints
* Generalise mmc async patch, running validation test is ongoing.
* Studied wl12xx_sdio_test, to be used for sdio async request validation.
=== Plans ===
* Get started with Snowball.
== Deepak Saxena <dsaxena> ==
=== Highlights ===
* Coordinating development of 11.11 blueprints
* Coordination of closing out 11.05 items:
- DT support for IGEP and u8500
- Started work on 11.05 kernel release
* Sent 11.11 Technical Requirements change request to TSC
* 1:1 with Jason H., John S., Mounir B.
=== Plans for Next Week ===
* Final review of blueprints and work items for 11.11 cycle
* Develop presentation for planning review (Due 05/31 to TSC)
* Put together outline for "ARM Porting Guide"
* Start scheduling regular 1:1s with team members
* Meeting with Zach, re: Android team needs, driver best practices howto
* Meeting with Paul L., re: Kernel testing process
* Meeting with Paul M., re: KWG 11.11 cycle
* Get Natty installed and up and running on new work laptop
=== Upcoming Travel/Time Off ===
Out Friday June 3rd, possibly on IRC and email
Out Friday June 17th - Mon June 20th
=== Highlights ===
* Posix Alarm Timers merged in mainline for 2.6.40-rc1
* Number of clocksource cleanups merged for 2.6.40-rc1
* A number of my RTC fixes were merged for 2.6.39
* My re-implementation of the virtual-battery driver functionality into
test_power driver is being pushed by the maintainer for 2.6.40
* task->comm locking fixes shot down by Ingo/Linus
* Sadly this burned a bit of time.
* But ok, as I wasn't strongly motivated to push the patches.
* Updated Linaro+Android git tree
* Included recent HDMI fixes from nico's tree
* Met with Deepak for 1:1 lunch.
* Took an initial swipe at blueprint workitems for 11.11 cycle.
* Emailed with Rafael and spent more time looking at pm_stay_awake
infrastructure
* Reviewed and tested some clocksource patches for Thomas which will
help extend NOHZ idle times beyond 5 seconds.
=== Plans ===
* Rebase and freshen up my current development git trees
* Continue researching pm_stay_awake/pm_relax infrastructure
* Try to adapt Arve's earlier RTC suggestions into suspend/resume path
=== Issues ===
* NA
Hi all,
Since we didn't have any meetings last two weeks, I wanted to send out
a reminder that we are having
our regular 8am PST and 8pm PST meetings this Monday. #linaro-kernel
on freenode.
Thanks,
~Deepak
== Thomas Abraham <thomas-ab> ==
=== Highlights ===
* Listed the work items for memory regions TR for
linaro-kernel-o-general-performance blueprint.
* Submitted patches of clkdev support on Samsung platforms. This is
prerequisite for common clock and device tree support.
(http://www.mail-archive.com/linux-samsung-soc@vger.kernel.org/msg04708.html)
* Continued testing of kexec-crashdump work item on s5pv310 platform.
Could not identify the reason for kernel reboot failure. Will continue
to debug this.
=== Plans ===
* Complete testing of kexec-crashdump on s5pv310 platform.
* Submit first version of common clock support for Samsung platforms.
Hi all,
We're trying to close out on a few 11.05 items and two that stand out
are completing DT support on IGEP and u8500 boards. The code is in
place in the linaro-2.6.38 kernel but we need someone with HW to debug
why it's not working. I'm working on sorting out the u8500 testing,
but does anyone in the kernel team or landing team have the HW on hand
and able to coordinate with Grant on debugging on the IGEP?
~Deepak
