#
989ee4e8 |
|
13-Jun-2018 |
David Moore <davemoore@google.com> |
[syscalls] Change vmar_... calls - Create zx_vm_option_t type - Change ZX_VM_FLAG_... to ZX_VM_... and make them new type. - Change int32_t flags parameters to zx_vm_option_t options - Make the options parameter be second. - Modify vmar_..._old functions to keep other repos working. Once they've all been updated these functions will go away. ZX-2264 Test:CQ Change-Id: I1faf4cc4e1e4dabf3d8ef680e398d83a545b0f09
|
#
0f9088d6 |
|
16-Jul-2018 |
John Grossman <johngro@google.com> |
[vmo-utils] Move PinnedVmo from IntelHDA utils into libfzl Promote the PinnedVmo class out of the IntelHDA driver and up to libfzl so that it can be used by other drivers. ZX-2462 #comment PinnedVmo class promoted into libfzl from HDA driver Test: Build and manual test with existing HDA driver. Change-Id: I60b2501c906e5ef3a58e9d6f1dc4ded6cfa3a191
|
#
beed9a1c |
|
10-Aug-2018 |
Yvonne Yip <yky@google.com> |
[dev][audio][intel-hda][dsp] shutdown the dsp in suspend() If the DSP is active while the system is powered off, it will cause the device to hang during shutdown because it is unable to go into S5. Powering off the DSP during shutdown resolves this issue. ZX-2185 #comment part of the fix TEST=manual testing using "dm poweroff" Change-Id: Ic2eea78fc04063d38f8978d7cdb2694a93847006
|
#
b33437b1 |
|
26-Jul-2018 |
Nick Maniscalco <maniscalco@google.com> |
[zircon][kernel] Use proper time types and overflow-safe arithmetic This change is preparation for converting zx_time_t and zx_duration_t from unsigned to signed types (ZX-2100). 1. Fix several places where the wrong time type was used (e.g. zx_time_t instead of zx_duration_t). Because these are simple typedefs it's easy to use the wrong one. While it generally doesn't cause a problem, using the wrong type can be misleading and lead to bugs like passing a relative time when an absolute time is expected. 2. Update kernel/ and some other parts of zircon to use the new overflow-safe arithmetic functions from zircon/time.h. Performance: Ran "/pkgfs/packages/zircon_benchmarks/0/bin/app -p" on NUC before and after this change. Differences were withing the noise, some slightly faster, some slightly slower. Test: CQ ZX-2100 #comment cleanup before conversion Change-Id: I75d074e5850ade5385decd9bbf330497bf7610e9
|
#
ff356b8b |
|
09-Aug-2018 |
John Grossman <johngro@google.com> |
[vmo-utils][fzl] Move contents of vmo-utils into fzl Move copies of the vmo-centric helper classes which existed in vmo-utils into fzl. This is step one of merging vmo-utils and libfzl. Once higher level code has been migrated over to use the new library, vmo-utils and its tests can be removed. Also, merge the libfzl and fzl unit tests. I'm not sure how they got divided in the first place, but now they are unified under just libfzl. ZX-2462 #comment Classes have been duplicated and added to libfzl Test: Build and unit-test Change-Id: Ie3bc07a9f78b6383032fcc974f28dd6af50caf21
|
#
24b85afe |
|
20-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] check FW_STATUS first during FW init After starting DMA for DSP firmware load, poll the FW_STATUS field to equal ADSP_FW_STATUS_STATE_ENTER_BASE_FW before waiting for the FW Ready IPC. Both conditions must be true when the DSP is operational. It seems like reading FW_STATUS has some effect. Previously the driver waited for the FW Ready IPC first, and sometimes the wait times out with a bad FW_STATUS. If we then poll FW_STATUS, it transitions to ADSP_FW_STATUS_STATE_ENTER_BASE_FW. Also stop the DMA after polling FW_STATUS and before waiting for IPC. This doesn't seem to have an effect. Test: booted a pixelbook that previously failed at least 1 out of 10 boots, now booted successfully 30 times. booted a second pixelbook 10 times successfully ZX-1538 #comment another potential fix Change-Id: I86989291e7e376f871beea3a0224e8d95323750e
|
#
ff28108c |
|
07-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] adjust log levels Change-Id: I1543e7a3175f5b736945649637c357e5fb5ff909
|
#
219c5763 |
|
14-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] print FW_STATUS on fw load failure Change-Id: I781d97e09c0fe95235fecd406655dad331f58cce
|
#
2eba1fc6 |
|
14-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] keep FW pinned until it is loaded The FW VMO was getting unpinned while the code loader DMA was in progress, because the PinnedVmo object was destroyed on function return from IntelDspCodeLoader::TransferFirmware() which is not responsible for DMA completion wait. Move the PinnedVmo object to the function that waits for firmware load success (implies DMA completion), so the memory stays pinned. Change-Id: I207a7ca6a68a448011eea6829251c2040d29d986
|
#
66dda880 |
|
05-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] set cldma size to stripped fw size Change-Id: I5421ca32740ab04b3b415c2c20daaae15c5abd36
|
#
21cadd74 |
|
05-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] poll the CPA bit for power up/down The driver sets the SPA bit to power up/down the DSP cores and polls the CPA bit to match the SPA value. When the values match the power transition has been achieved. Previously it was polling the wrong bit so the DSP may be in the wrong state for subsequent operations. Change-Id: I6ce9a3fa7bfcf9978b443da732e03da33e0201f2
|
#
8f4fe9b3 |
|
04-Jun-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] use FW_READY ipc notification Wait for both the FW_READY IPC notification and fw_status when waiting for the DSP to load firmware. Change-Id: I35edb74f31ab20e3756bce529f4b7f1b52dd782b
|
#
b0826017 |
|
22-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] pass dsp pipeline param to stream Change-Id: I52605820c06dddc0b41a2e75e9ed8e878b32146f
|
#
35fa2ff3 |
|
22-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] introduce ipc workqueue Change-Id: Ia1aeefd92fae121341a0974b005b0167d91da154
|
#
a57932fa |
|
18-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] get I2S config blob from NHLT Change-Id: I2afd4bc5d7a256636d64cd2d70a1531bad1cd897
|
#
e43b2d17 |
|
25-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] correct shutdown sequence Must disable IRQs, then stall/reset the cores before powering down. ZX-2185 #done Change-Id: I8783af42d5dd312765ef9f19b4d817e3a845cdf4
|
#
fb8930da |
|
14-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] add dsp modules for capture Change-Id: I99badae28a3c510201cf19f099dc8a094098f9c0
|
#
51f9947a |
|
10-May-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda][dsp] implement audio api Derive IntelAudioDsp from IntelHDACodecDriverBase and derive IntelDspStream from IntelHDAStreamBase to reuse HDA stream management code. The DSP streams maintains additional channels to intercept ring-buffer channel requests, so the DSP can coordinate starting and stopping pipelines with the ring-buffer. Change-Id: If935cbe4e97d0a940bf3c39b8ba6cfeb4c707438
|
#
ef61bf75 |
|
04-Apr-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-audio-dsp] initialize basic DSP pipeline Initialize a single pipeline for speaker playback. Change-Id: I7eea457e4a53a804942924b90c1bbf35b833a5f1
|
#
7d35c383 |
|
14-Feb-2018 |
Yvonne Yip <yky@google.com> |
[dev][intel-hda] intel-audio-dsp driver Currently just boots the DSP and loads the firmware. The DSP driver is independent from the HDA driver, however the source is checked in under system/dev/audio/intel-hda/ to be consistent with HDA codecs. Change-Id: Ic2d9bca68ae7ee6c9011f53d0f894b7a68e43b86
|