Aug 27 2018
Aug 24 2018
Aug 22 2018
This patchset has one caveat at this time that will be fixed soon. If we go into kernel space during the transaction, and the process is rescheduled to another cpu, the transaction state will be on the original CPU, and the new CPU will not have the same CPU state, caused an undefined behaviour.
Adding additional testsuite for the code.
Aug 7 2018
Jul 27 2018
Add aditional tests. Mainly two things:
- Instead of testing for transactional states, look for active (Transactional + Suspended)
- Avoid signal handling inside a transaction, since the code will be rolled back.
Jun 19 2018
Adding Justin's suggestions and removing an unused variable.
Jun 15 2018
Jun 12 2018
Fixing silly error.
Jun 11 2018
Adjusting to follow jhibbits' suggestion.
Better approach forcing promotion.
Jun 8 2018
Change OF_getencprop() to OF_hasprop() on probe() function.
Jun 7 2018
Jun 6 2018
Removing the volatile qualifier for mfs_root and mfs_root_size.
Jun 5 2018
I tested it on my POWER9 and it continues to boot. This change seems to make sense, and fixes 32-bits initrd addresses, but I am not sure how it broke Justin's P9, and if it is going to fix.
Jun 1 2018
May 30 2018
May 29 2018
May 28 2018
I tested this change on my POWER8 pseries system and it worked fine.
May 21 2018
I reviewed this code and it seems OK to be integrated.
May 18 2018
- Is there a document describing this that we can reference?
May 16 2018
Changes according to the Justin's review:
May 14 2018
May 11 2018
May 9 2018
Thanks Leandro. This closes bug#227093
May 7 2018
May 4 2018
Removing extra line.
Apr 30 2018
Apr 23 2018
Apr 18 2018
Apr 17 2018
Use SPR_DSCR macro instead of SPR number (0x11)
Removing mfdscr() inline function and replacing by mfspr(SPR_DSCR).
Apr 16 2018
Apr 6 2018
Updating the patch with mjg's suggestions. Thanks for the review.
Apr 5 2018
Updating the patch to remove the mem alloc from outside the lock section.
Apr 3 2018
Mar 7 2018
This is working fine on pseries also. The problem I reported above was found to not be related to this revision.
Mar 6 2018
This is fascinating. Could you verify that this happens without this patch (I assume it is unrelated?) and give some more details on your setup that gave this error. I would like to try to get to the bottom of it.
This is the regular boot (without usefdt=1).
With the current patch, I was able to test it properly on my pseries VM, but I didn't use usefdt=1, mainly because I didn't know where to use it. Should I need to recompile the loader to use it, or, is it a parameter somewhere?
And it was working fine?
Mar 5 2018
Looking at DSISR (0x40000000), it seems that the only bit enable represents the following thing:
I tried you most recent patchset and it is still not working on pseries.
Mar 4 2018
Mar 3 2018
Patchset v2. Changes as requested by Justin (Thanks!):
hi Justin, thanks for the review. I will work on a v2 patchset.
Mar 2 2018
Feb 26 2018
If the above analyzes is correct, It seems that a function ends up with a conditional branch, so, if the condition is not reached, then it moved to the function epilogue, which is 0x0.
Once my machine stops, it just shows:
Resubmitting a code that treats the errors other than ignoring them,.
Feb 23 2018
Feb 9 2018
Feb 8 2018
Let me attach a new fix as proposed by Justin