Hi,
Linaro kernels include a patch by Tony Lindgren that adds a printascii() to every printk() call. Currently in linux-linaro-tracking, that's 892a9def (ARM: Make low-level printk work). With this patch, every line is printed twice once you've registered a console driver which is a bit of a pain. Speaking to him, he acknowledges that this hack is no longer needed since we now have earlyprintk, so can this patch be removed from linaro kernels?
Cheers, Javi
PS: Please CC me, I'm not subscribed to the list
In my tree I have this patch on top of llct to get rid of that nasty echo
diff --git a/kernel/printk.c b/kernel/printk.c index 1e75d13..97f950d 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -1547,10 +1547,6 @@ asmlinkage int vprintk_emit(int facility, int level, */ text_len = vscnprintf(text, sizeof(textbuf), fmt, args);
-#ifdef CONFIG_DEBUG_LL - printascii(text); -#endif - /* mark and strip a trailing newline */ if (text_len && text[text_len-1] == '\n') { text_len--;
On 13 June 2013 17:29, Javi Merino javi.merino@arm.com wrote:
Hi,
Linaro kernels include a patch by Tony Lindgren that adds a printascii() to every printk() call. Currently in linux-linaro-tracking, that's 892a9def (ARM: Make low-level printk work). With this patch, every line is printed twice once you've registered a console driver which is a bit of a pain. Speaking to him, he acknowledges that this hack is no longer needed since we now have earlyprintk, so can this patch be removed from linaro kernels?
Cheers, Javi
PS: Please CC me, I'm not subscribed to the list
linaro-kernel mailing list linaro-kernel@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-kernel
On Thu, 2013-06-13 at 10:29 +0100, Javi Merino wrote:
Hi,
Linaro kernels include a patch by Tony Lindgren that adds a printascii() to every printk() call. Currently in linux-linaro-tracking, that's 892a9def (ARM: Make low-level printk work). With this patch, every line is printed twice once you've registered a console driver which is a bit of a pain. Speaking to him, he acknowledges that this hack is no longer needed since we now have earlyprintk, so can this patch be removed from linaro kernels?
By the looks of things, it's already been removed, and was done so a couple of weeks ago before the Linaro 13.05 release. The current commit which reverts it is... https://git.linaro.org/gitweb?p=kernel/linux-linaro-tracking.git%3Ba=commit%...
On 06/13/2013 03:27 PM, Jon Medhurst (Tixy) wrote:
On Thu, 2013-06-13 at 10:29 +0100, Javi Merino wrote:
Hi,
Linaro kernels include a patch by Tony Lindgren that adds a printascii() to every printk() call. Currently in linux-linaro-tracking, that's 892a9def (ARM: Make low-level printk work).
Right. It has been re-added in the middle of 13.05 cycle when half of this patch (the 2nd of its two hunks) has got into one of the topics.
With this patch, every line is printed twice once you've registered a console driver which is a bit of a pain. Speaking to him, he acknowledges that this hack is no longer needed since we now have earlyprintk, so can this patch be removed from linaro kernels?
By the looks of things, it's already been removed, and was done so a couple of weeks ago before the Linaro 13.05 release. The current commit which reverts it is... https://git.linaro.org/gitweb?p=kernel/linux-linaro-tracking.git%3Ba=commit%...
I haven't noticed that the Android topic reverted this change. Now I've removed the 892a9def from llct-v3.10-misc-fixes topic. So the "Make low-level printk work" commit is now removed from linux-linaro-core-tracking, and will get removed from linux-linaro on its next update (should be tomorrow).
Thanks, Andrey
On 06/13/2013 02:29 AM, Javi Merino wrote:
Linaro kernels include a patch by Tony Lindgren that adds a printascii() to every printk() call. Currently in linux-linaro-tracking, that's 892a9def (ARM: Make low-level printk work). With this patch, every line is printed twice once you've registered a console driver which is a bit of a pain. Speaking to him, he acknowledges that this hack is no longer needed since we now have earlyprintk, so can this patch be removed from linaro kernels?
Yea. As Andy mentioned, we carry a patch to revert that change in the Linaro tree (though occasionally I get forgetful and it gets dropped when migrating to new AOSP releases).
I tried to push the revert w/ Andy's rational to AOSP but the Android devs apparently have a fondness for the patch, and still use it: https://android-review.googlesource.com/#/c/59074/
(My guess is they consider it easier to just keep the patch then to have to remember the boot arguments required when debugging difficult problems.)
thanks -john
linaro-kernel@lists.linaro.org
-
Andrey Konovalov -
Andy Green -
Javi Merino -
John Stultz -
Jon Medhurst (Tixy)