h1. PeriPeri-2019-02

| Dates      | 2019-02-02 - 2019-02-03 |         |
| Place      | see 'Activities' below |          |
|/10. Member | Geert   | 2019-02-01 - 2019-02-03 |
             | Jacopo  | 2019-01-31 - 2019-02-05+|
             | Kieran  | 2019-02-01 - 2019-02-05+|
             | Laurent | 2019-02-03 - 2019-02-05+|
             | Magnus  | mostly 2019-02-02 + 2019-02-03 |
             | Marek   | (wants to attend Fri+Sun meetings) |
             | Niklas  | 2019-01-31 - 2019-02-05+|
             | Simon   | 2019-01-31 - 2019-02-04 |
             | Ulrich  | 2019-01-31 - 2019-02-04 |
             | Wolfram | 2019-02-01 - 2019-02-05 |

h2. Activities in Brussels

| Date                 | Time              | Activity                                 | Location                                  |
| Friday 2019-02-01    | 09:00 - 17:00     | Core, I/O Group                          | http://www.bbl2meet.be/ (Penthouse)       |
| Friday 2019-02-01    | 19:00             | Dinner                                   | http://www.kokob.be/                      |
| Saturday 2019-02-02  | 10:00 - 19:00     | FOSDEM                                   | https://fosdem.org/2019/                  |
| Sunday 2019-02-03    | 09:00 - 17:00     | FOSDEM                                   | https://fosdem.org/2019/                  |
| Sunday 2019-02-03    | 19:00 - ?         | Global Dinner Meeting                    | https://www.winebarsablon.be/eng/ (Salon) |
| Tuesday 2019-02-05   | 09:00 - 12:00     | MM Group                                 | ?                                         |

h2. Hardware wishlist

| Hardware             | Requested by      | Brought by        | Comment                             |
| Ebisu board          | Wolfram           | Geert             | IIC_DVFS testing on Gen3 (hopefully an hour will do) |

!2019-02-periperi/fosdem.png!

h2. Agenda for Friday Meeting (Core + IO)

h3. Core

* ATF:
** Document installation procedure at elinux.org wiki
** Configure from DTS
** UBoot and ATF could be built through CI and tested.
*** e.g. https://travis-ci.org/marex/arm-trusted-firmware-1/builds/485750085
*** https://docs.travis-ci.com/user/uploading-artifacts/ for prebuilt binaries ?

* U-Boot: MMC HS200/HS400 and SD SDR104 issues.
** Playing whack-a-mole with bug fixes.
** Testing identified bugs in core U-Boot MMC code which is being explored and fixed.

* OpenOCD
** Kernel awareness patches for GDB are progressing *SLOWLY*.
** A Latest RFC is posted by Linaro:
*** https://sourceware.org/ml/gdb-patches/2019-01/msg00596.html

* BSP
** List is reducing for two reasons
*** We've got stuff done
*** BSP is filtered to priorities.

* Upstream
  Maintainers are happy. Renesas is seen as the 'second' best upstreaming ARM-SoC.
  The first is ATMEL which is so small it's incomparable - they have less work to do :)

* CPUidle:
** Get clarification of real issues (SYSC CPU/GSX concurrency?)
** Enable when we get newer SoC revisions where the hardware has been fixed

* IPMMU
** Unusable on H3 before ES3.x..
** E3/M3 might be OK (because they are newer)
** 32-bits IOVA space status? Discuss at Sundays IPMMU chat

* TEE
** Coordinate between optee os and Linux -> Marek?

* 32-bit DMA limitation on arm64 fixed upstream?

* Gen3 TDSEL:
** Assumption: (1) reset state or (2) value programmed by bootloader is OK
** (2) => Add register to list of registers to save/restore during suspend/resume
** Describe TDSEL value in DT?

* Patchwork and automatic status updates.
** Patchwork now has automatic filters to delegate to Geert / Horms / Kbingham
** A few false positives on the rules, but otherwise helpful.
** New rules to catch missing corners.
** Horms to suggest any extra rules to match maintained files
** KB to request automatic patchwork update tool against linux-next.


h3. IO

* plan next tasks
* sync SDHI experiences (esp. Linux / U-Boot)
* [[upstream design for complex GMSLish I2C setups]]

http://muistio.tieke.fi/p/PeriPeriBE

h2. Agenda for Sunday Meeting (global & casual)

* IPMMU (moved to Global due to Magnus' availability)
** What is the performance impact of enabling the IPMMU?
*** Needs benchmarking
*** IPMMU may not be useful for boards with less than 4 GiB of RAM

* Periject
# Missing feature is automatic ID generation
*** To be implemented (on client)
# Run scripts to populate database from older data
# Go life
# Profit! ;-)

* Discussion about future plans
** R-Car H3-N: No new information
** Virtualization: No feedback from Renesas

h2. Agenda for Tuesday Meeting (MM)


h1. Files

* "pdf: Core + Global":../../wiki/2019-02-periperi/PERIPERI_BE.pdf
* "pdf: IO":../../wiki/2019-02-periperi/PeriPeriBE-State_of_the_IO.pdf
* "jpg: Sunday Meeting 1":../../wiki/2019-02-periperi/195838.jpg
* "jpg: Sunday Meeting 2":../../wiki/2019-02-periperi/195912.jpg