On 2019-08-22 22:55, Pierre-Louis Bossart wrote:
On 8/22/19 2:03 PM, Cezary Rojewski wrote:
This segment targets initialization procedures for all SST drivers. For legacy:
- it concentrates on updating ACPI loader so generic DSP framework is
not fed any platform specific data. sst-acpi module ceases to exist and is replaced by hsw-acpi and byt-acpi
For cAVS (Skylake+):
- Repairs what is currently an initialization mess, given the order of
invocation of engaged handlers and sheer amount of them.
- provides interface to offload hardware-specifics away from driver
Following is the order of Skylake initialization currently:
- skl_probe
-> schedule skl_probe_work
- skl_first_init
- skl_create
-> skl_init_dsp --> skl/ bxt/ cnl_sst_dsp_init ---> skl_sst_ctx_init ----> skl_dsp_ctx_init -----> sst_dsp_new ------> sst_ops::init (invoked but unimplemented!)
Listing all the types engaged together with the naming chosen for the above paints an even darker picture. Code is unreadable and hides initialization for diffenret members is various handlers. Moreover, due to existing ill relationship between skl_dev, sst_dsp and sst_generic_ipc, one must be extra careful when accessing so called "dsp/ thread_context" as they all don't get initialized immediately before object is yield for further processing. However, 100/100 series is nonsense and thus cleanups have been divided into chunks and prioritized.
Code seen here is part of new Skylake fundament, located at the very bottom of internal mainline. Said mainline is tested constantly on at least sigle platform from every cAVS bucket (description below). This week, BDW has been added to the CI family and was essential in validating legacy changes. Baytrail platform is still missing. Changes for BYT directly mirror HSW/ BDW but due to current lack of platform were untested. Boards engaged in testing: rt286, rt298, rt274.
this is not enough, sorry. these are RVPs and you need to check with commercial devices supported in sound/soc/intel/boards/.
+Filip +Michal
What machine board has to do with FW and host side? If it has, we better notify the owner so he can fix codec's code at once. All boards MUST follow recommended protocol whether its HDA or I2S in communicating with /skylake. This is hardware IP we taking about. I could as well test all platforms with AudioPrecision and say: shipit.
DSP "commercial devices" with 99% of home audio being routed through HD-Audio legacy? I do contact representatives of "commercial devices" daily, you of all should be aware of fact that in almost all cases they are fed neither with upstream code nor upstream binaries. For the first time since eons sound/soc/intel/skylake code is tested before upstreaming, yet you still defend the mistakes of the past?
I don't mind being extra careful, but testing every existing machine board is nonsense and you know it. Extra careful: this should have been the main topic enforced by Intel audio representatives when existing stuff been upstreamed - years ago.
!!! IMPORTANT !!!
Some upstream FW binaries are not compatible with existing /skylake driver while changes found here (HARDWARE_CONFIG/ FIRMWARE_CONFIG) make use of firmware ability to offload hardware-specifics away from driver. These and more are core part of any cAVS design and are to be implemented and used by host. This too is missing on Linux upstream.
As explanined once, five main FW branches are available:
- kbl, 1.5 cAVS (supports SKL, KBL, KBL-R, ABL, more..)
- apl_auto, 1.5+ cAVS (supports APL, GLK)
- cnl, 1.8 cAVS (supports CNL, CFL, WHL, CML)
- icl, 2.0 cAVS (supports ICL, LKF)
- master, 2.5 cAVS (supports TGL, EHL and everything above)
SKL FW binary existing on upstream is a descendant of old spt branch, obsoleted for 4-5 years now. That FW is a stub, quickly replaced by kbl which is to be used on all 1.5 cAVS platforms.
The same story goes with bxtn binary, a descendant of apl branch, replaced by apl_auto 2-3 years ago. All vendors, entire validation and development is located on apl_auto.
Message: all FW binaries are to be updated as we cannot guarantee these are still functional. Given the fact that all vendors are fed with new binaries on regular basis from all main branches, it's highly probable some scenarios fail with existing FWs. In consequence, linux-firmware patch will be provided with fresh, updated binaries, soon. Once more, please note, we do not support, nor test platforms using obsolete FW binaries. That includes all platforms, not just SKL or APL.
"we don't break userspace" means you have to be backwards compatible with whatever the user has on their filesystem when the kernel is updated. Or if you go ahead, you need to detect that the firmware is not supported and make it known to the user that it needs to be updated.
Gotta be honest here. I'm surprised SPT binary even made it here. Same goes for bxtn one. The only real-life example of bxtn for linux is for 4.1.X kernels - where YES, /skylake and topology and a lot more stuff does not even exist and is simply backported internally.
If I could, I'd rather prefer the "detect and notify" as it is impossible to repair all the mistakes made in /sound/soc/intel/skylake.
However, in practice there isn't any reliable way to verify the actual usability of old FW binary against host site as the interface is volatile and numbers alone don't mean much. Patch with FW binaries would not remove old ones, simply add new versions to the directory.
Czarek
Amadeusz Sławiński (1): ASoC: Intel: Skylake: Put FW runtime params defs in one place
Cezary Rojewski (34): ASoC: Intel: Skylake: Add FIRMWARE_CONFIG IPC request ASoC: Intel: Skylake: Add HARDWARE_CONFIG IPC request ASoC: Intel: Skylake: Unify firmware loading mechanism ASoC: Intel: Skylake: Reload libraries on D0 entry for CNL ASoC: Intel: Skylake: Unhardcode dsp cores number ASoC: Intel: Skylake: Update interrupt disabling routine ASoC: Intel: Skylake: Inline ipc free operations ASoC: Intel: Skylake: Unify driver cleanup mechanism ASoC: Intel: Relocate irq thread header to sst_ops ASoC: Intel: Merge sst_dsp_device into sst_pdata ASoC: Intel: Skylake: Reuse sst_dsp_free ASoC: Intel: Skylake: Reuse sst_dsp_new ASoC: Intel: Skylake: Remove skl_dsp_acquire_irq ASoC: Intel: Skylake: Use dsp loading functions directly ASoC: Intel: Skylake: Make dsp_ops::stream_tag obsolete ASoC: Intel: Skylake: Remove skl_dsp_loader_ops ASoC: Intel: Skylake: Remove window0 sst_addr fields ASoC: Intel: Skylake: Remove redundant W0 and W1 macros ASoC: Intel: Skylake: Remove redundant SRAM fields ASoC: Intel: Expose ACPI loading members ASoC: Intel: Haswell: Define separate ACPI loader ASoC: Intel: Baytrail: Define separate ACPI loader ASoC: Intel: Refactor probing of ACPI devices ASoC: Intel: Skylake: Simplify skl_sst_ctx_init declaration ASoC: Intel: Skylake: Simplify all sst_dsp_init declarations ASoC: Intel: Skylake: Define platform descriptors ASoC: Intel: Skylake: Update skl_ids table ASoC: Intel: Skylake: Flip SST initialization order ASoC: Intel: Reuse sst_pdata::fw_name field ASoC: Intel: Reuse sst_pdata::fw field ASoC: Intel: Skylake: Remove skl_dsp_ops ASoC: Intel: Skylake: Privatize SST init handlers ASoC: Intel: Skylake: Merge skl_sst_ctx_init into skl_init_dsp ASoC: Intel: Remove obsolete firmware fields
sound/soc/intel/Kconfig | 14 +- sound/soc/intel/baytrail/Makefile | 2 + sound/soc/intel/baytrail/acpi.c | 64 +++++ sound/soc/intel/baytrail/sst-baytrail-dsp.c | 2 +- sound/soc/intel/baytrail/sst-baytrail-ipc.c | 13 +- sound/soc/intel/baytrail/sst-baytrail-ipc.h | 2 + sound/soc/intel/common/Makefile | 4 +- .../intel/common/soc-acpi-intel-bxt-match.c | 2 - .../intel/common/soc-acpi-intel-byt-match.c | 2 - .../intel/common/soc-acpi-intel-cnl-match.c | 1 - .../intel/common/soc-acpi-intel-glk-match.c | 3 - .../intel/common/soc-acpi-intel-hda-match.c | 2 - .../common/soc-acpi-intel-hsw-bdw-match.c | 4 - .../intel/common/soc-acpi-intel-icl-match.c | 1 - .../intel/common/soc-acpi-intel-kbl-match.c | 12 - .../intel/common/soc-acpi-intel-skl-match.c | 3 - sound/soc/intel/common/sst-acpi.c | 117 +-------- sound/soc/intel/common/sst-dsp-priv.h | 8 +- sound/soc/intel/common/sst-dsp.h | 37 +-- sound/soc/intel/common/sst-firmware.c | 13 +- sound/soc/intel/haswell/Makefile | 2 + sound/soc/intel/haswell/acpi.c | 78 ++++++ sound/soc/intel/haswell/sst-haswell-dsp.c | 1 + sound/soc/intel/haswell/sst-haswell-ipc.c | 13 +- sound/soc/intel/haswell/sst-haswell-ipc.h | 2 + sound/soc/intel/skylake/bxt-sst.c | 143 ++++------- sound/soc/intel/skylake/cnl-sst-dsp.c | 13 +- sound/soc/intel/skylake/cnl-sst-dsp.h | 15 +- sound/soc/intel/skylake/cnl-sst.c | 142 ++++------- sound/soc/intel/skylake/skl-debug.c | 2 +- sound/soc/intel/skylake/skl-messages.c | 194 ++------------ sound/soc/intel/skylake/skl-pcm.c | 22 +- sound/soc/intel/skylake/skl-sst-cldma.c | 10 +- sound/soc/intel/skylake/skl-sst-dsp.c | 79 ++---- sound/soc/intel/skylake/skl-sst-dsp.h | 55 ++-- sound/soc/intel/skylake/skl-sst-ipc.c | 238 ++++++++++++++++-- sound/soc/intel/skylake/skl-sst-ipc.h | 124 ++++++++- sound/soc/intel/skylake/skl-sst-utils.c | 28 +-- sound/soc/intel/skylake/skl-sst.c | 150 ++++++----- sound/soc/intel/skylake/skl.c | 65 +++-- sound/soc/intel/skylake/skl.h | 26 +- 41 files changed, 877 insertions(+), 831 deletions(-) create mode 100644 sound/soc/intel/baytrail/acpi.c create mode 100644 sound/soc/intel/haswell/acpi.c