Kevin Chiu has uploaded this change for review. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: Add variant.c to align with dalboz reference board ......................................................................
mb/google/zork: Add variant.c to align with dalboz reference board
BUG=b:161579679 BRANCH=master TEST=1. emerge-zork coreboot chromeos-bootimage 2. power on proto board successfully
Change-Id: I9dffdf5654680e3c2c0b259ee82a471f8ff14f56 Signed-off-by: Kevin Chiu kevin.chiu@quantatw.com --- M src/mainboard/google/zork/variants/dirinboz/Makefile.inc A src/mainboard/google/zork/variants/dirinboz/variant.c 2 files changed, 214 insertions(+), 0 deletions(-)
git pull ssh://review.coreboot.org:29418/coreboot refs/changes/43/44343/1
diff --git a/src/mainboard/google/zork/variants/dirinboz/Makefile.inc b/src/mainboard/google/zork/variants/dirinboz/Makefile.inc index 57e7136..51d19fe 100644 --- a/src/mainboard/google/zork/variants/dirinboz/Makefile.inc +++ b/src/mainboard/google/zork/variants/dirinboz/Makefile.inc @@ -3,3 +3,4 @@ subdirs-y += ./spd
ramstage-y += gpio.c +ramstage-y += variant.c diff --git a/src/mainboard/google/zork/variants/dirinboz/variant.c b/src/mainboard/google/zork/variants/dirinboz/variant.c new file mode 100644 index 0000000..de4338e --- /dev/null +++ b/src/mainboard/google/zork/variants/dirinboz/variant.c @@ -0,0 +1,213 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +#include <baseboard/variants.h> +#include <console/console.h> +#include <device/device.h> +#include <drivers/i2c/generic/chip.h> +#include <soc/pci_devs.h> +#include <ec/google/chromeec/ec.h> +#include <ec/google/chromeec/i2c_tunnel/chip.h> +#include <string.h> + +#define EC_PNP_ID 0x0c09 + +/* Look for an EC device of type PNP with id 0x0c09 */ +static bool match_ec_dev(DEVTREE_CONST struct device *dev) +{ + if (dev->path.type != DEVICE_PATH_PNP) + return false; + + if (dev->path.pnp.port != EC_PNP_ID) + return false; + + return true; +} + +extern struct chip_operations drivers_i2c_generic_ops; + +/* Look for an I2C device with HID "10EC5682" */ +static bool match_audio_dev(DEVTREE_CONST struct device *dev) +{ + struct drivers_i2c_generic_config *cfg; + + if (dev->chip_ops != &drivers_i2c_generic_ops) + return false; + + cfg = dev->chip_info; + + return !strcmp(cfg->hid, "10EC5682"); +} + +extern struct chip_operations ec_google_chromeec_i2c_tunnel_ops; + +/* Look for Cros EC tunnel device which has audio device under it. */ +static bool match_audio_tunnel(DEVTREE_CONST struct device *dev) +{ + const struct device *audio_dev; + + if (dev->chip_ops != &ec_google_chromeec_i2c_tunnel_ops) + return false; + + audio_dev = dev_find_matching_device_on_bus(dev->link_list, match_audio_dev); + + if (!audio_dev) + return false; + + return true; +} + +/* + * This is to allow support for audio on older board versions (< 2). [b/153458561]. This + * should be removed once these boards are phased out. + */ +static void update_audio_configuration(void) +{ + uint32_t board_version; + const struct device *lpc_controller; + const struct device *ec_dev; + const struct device *i2c_tunnel_dev; + struct ec_google_chromeec_i2c_tunnel_config *cfg; + + /* If CBI board version cannot be read, assume this is an older revision of hardware. */ + if (google_chromeec_cbi_get_board_version(&board_version) != 0) + board_version = 1; + + if (board_version >= 2) + return; + + lpc_controller = SOC_LPC_DEV; + if (lpc_controller == NULL) { + printk(BIOS_ERR, "%s: LPC controller device not found!\n", __func__); + return; + } + + ec_dev = dev_find_matching_device_on_bus(lpc_controller->link_list, match_ec_dev); + + if (ec_dev == NULL) { + printk(BIOS_ERR, "%s: EC device not found!\n", __func__); + return; + } + + i2c_tunnel_dev = dev_find_matching_device_on_bus(ec_dev->link_list, match_audio_tunnel); + + if (i2c_tunnel_dev == NULL) { + printk(BIOS_ERR, "%s: I2C tunnel device not found!\n", __func__); + return; + } + + cfg = i2c_tunnel_dev->chip_info; + if (cfg == NULL) { + printk(BIOS_ERR, "%s: I2C tunnel device config not found!\n", __func__); + return; + } + + cfg->remote_bus = 5; +} + +static int sku_has_emmc(void) +{ + uint32_t board_sku = sku_id(); + + /* Factory flow requires all OS boot media to be enabled. */ + if (boot_is_factory_unprovisioned()) + return 1; + + /* FIXME: This needs to be fw_config controlled. */ + /* Enable emmc0 for unknown skus. Only sku3/0xC really has it. */ + if (board_sku == 0x5A80000C || board_sku == 0x5A800003 || board_sku == CROS_SKU_UNKNOWN) + return 1; + + return 0; +} + +void variant_devtree_update(void) +{ + struct soc_amd_picasso_config *cfg; + + cfg = config_of_soc(); + + if (sku_has_emmc()) { + if ((sku_id() == 0x5A800003) || (sku_id() == 0x5A80000C)) { + /* + * rev0 and rev1 boards have issues with HS400 + * + * There is a tuning fix with ES which shows promise + * for some boards, and a HW fix with stitching vias. + * There were also concerns that these boards did not + * have good margins for certain skus. + * + * But these original boards have none of these fixes. + * So we keep the speed low here, with the intent that + * other variants implement these corrections. + */ + cfg->sd_emmc_config = SD_EMMC_EMMC_HS200; + } + } else { + cfg->sd_emmc_config = SD_EMMC_DISABLE; + } + + update_audio_configuration(); +} + +/* FIXME: Comments seem to suggest these are not entirely correct. */ +static const fsp_ddi_descriptor non_hdmi_ddi_descriptors[] = { + { + // DDI0, DP0, eDP + .connector_type = EDP, + .aux_index = AUX1, + .hdp_index = HDP1 + }, + { + // DDI1, DP1, DB OPT2 USB-C1 / DB OPT3 MST hub + .connector_type = DP, + .aux_index = AUX2, + .hdp_index = HDP2 + }, + { + // DP2 pins not connected on Dali + // DDI2, DP3, USB-C0 + .connector_type = DP, + .aux_index = AUX4, + .hdp_index = HDP4, + } +}; + +static const fsp_ddi_descriptor hdmi_ddi_descriptors[] = { + { // DDI0, DP0, eDP + .connector_type = EDP, + .aux_index = AUX1, + .hdp_index = HDP1 + }, + { // DDI1, DP1, DB OPT2 USB-C1 / DB OPT3 MST hub + .connector_type = HDMI, + .aux_index = AUX2, + .hdp_index = HDP2 + }, + // DP2 pins not connected on Dali + { // DDI2, DP3, USB-C0 + .connector_type = DP, + .aux_index = AUX4, + .hdp_index = HDP4, + } +}; + +void variant_get_dxio_ddi_descriptors(const fsp_dxio_descriptor **dxio_descs, + size_t *dxio_num, + const fsp_ddi_descriptor **ddi_descs, + size_t *ddi_num) +{ + uint32_t board_sku = sku_id(); + + *dxio_descs = baseboard_get_dxio_descriptors(dxio_num); + + /* SKU 1, A, and D DB have HDMI, as well as unknown */ + /* FIXME: this needs to be fw_config controlled. */ + if ((board_sku == 0x5A80000A) || (board_sku == 0x5A80000D) || (board_sku == 0x5A800001) + || (board_sku == CROS_SKU_UNKNOWN)) { + *ddi_descs = &hdmi_ddi_descriptors[0]; + *ddi_num = ARRAY_SIZE(hdmi_ddi_descriptors); + } else { + *ddi_descs = &non_hdmi_ddi_descriptors[0]; + *ddi_num = ARRAY_SIZE(non_hdmi_ddi_descriptors); + } +}
Aaron Durbin has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: Add variant.c to align with dalboz reference board ......................................................................
Patch Set 1:
Kevin, do you need all the logic in variant.c? Or was this copy/pasted?
Kevin Chiu has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: Add variant.c to align with dalboz reference board ......................................................................
Patch Set 1:
Patch Set 1:
Kevin, do you need all the logic in variant.c? Or was this copy/pasted?
Hi Aaron, it was just copied from dalboz as the initialization. thanks.
Furquan Shaikh has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: Add variant.c to align with dalboz reference board ......................................................................
Patch Set 1:
(4 comments)
I had added comments this morning, but did not hit submit before.
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG@7 PS1, Line 7: Add variant.c to align with dalboz reference board What is the motivation behind doing this? Some of the changes in variant.c for dalboz are used for supporting older revisions of the board (i.e. older schematics) which I don't think apply to dirinboz.
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... File src/mainboard/google/zork/variants/dirinboz/variant.c:
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 60: audio on older board versions I don't think this applies to dirinboz. Entire code from line 12-line 105 should be dropped.
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 117: board_sku == 0x5A80000C || board_sku == 0x5A800003 || board_sku == CROS_SKU_UNKNOWN Zork is moving away from using SKU ID to FW_CONFIG. Is this really correct for dirinboz?
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 205: (board_sku == 0x5A80000A) || (board_sku == 0x5A80000D) || (board_sku == 0x5A800001) All these SKU IDs are specific to Dalboz. Also, zork is moving to using FW_CONFIG instead of SKU IDs.
Hello build bot (Jenkins), Furquan Shaikh, Patrick Georgi, Martin Roth, Aaron Durbin,
I'd like you to reexamine a change. Please visit
https://review.coreboot.org/c/coreboot/+/44343
to look at the new patch set (#2).
Change subject: mb/google/zork: config ddi for dirinboz ......................................................................
mb/google/zork: config ddi for dirinboz
dirinboz does not support native HDMI, config DDI as below: DDI0: eDP DDI1: DP DDI2: DP
BUG=b:161579679 BRANCH=master TEST=1. emerge-zork coreboot chromeos-bootimage 2. power on proto board successfully
Change-Id: I9dffdf5654680e3c2c0b259ee82a471f8ff14f56 Signed-off-by: Kevin Chiu kevin.chiu@quantatw.com --- M src/mainboard/google/zork/variants/dirinboz/Makefile.inc A src/mainboard/google/zork/variants/dirinboz/variant.c 2 files changed, 45 insertions(+), 0 deletions(-)
git pull ssh://review.coreboot.org:29418/coreboot refs/changes/43/44343/2
Kevin Chiu has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: config ddi for dirinboz ......................................................................
Patch Set 2:
(4 comments)
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG@7 PS1, Line 7: Add variant.c to align with dalboz reference board
What is the motivation behind doing this? Some of the changes in variant. […]
Hi Furquan, thank you for the comment. it looks we only need to config DDI setting for dirinboz.
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... File src/mainboard/google/zork/variants/dirinboz/variant.c:
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 60: audio on older board versions
I don't think this applies to dirinboz. Entire code from line 12-line 105 should be dropped.
Done
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 117: board_sku == 0x5A80000C || board_sku == 0x5A800003 || board_sku == CROS_SKU_UNKNOWN
Zork is moving away from using SKU ID to FW_CONFIG. […]
removed, done.
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 205: (board_sku == 0x5A80000A) || (board_sku == 0x5A80000D) || (board_sku == 0x5A800001)
All these SKU IDs are specific to Dalboz. […]
removed, done. only config DDI to non HDMI for dirinboz.
Aaron Durbin has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: config ddi for dirinboz ......................................................................
Patch Set 2: Code-Review+2
Aaron Durbin has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: config ddi for dirinboz ......................................................................
Patch Set 2:
(3 comments)
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/coreboot/+/44343/1//COMMIT_MSG@7 PS1, Line 7: Add variant.c to align with dalboz reference board
Hi Furquan, thank you for the comment. it looks we only need to config DDI setting for dirinboz.
Ack
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... File src/mainboard/google/zork/variants/dirinboz/variant.c:
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 117: board_sku == 0x5A80000C || board_sku == 0x5A800003 || board_sku == CROS_SKU_UNKNOWN
removed, done.
Done
https://review.coreboot.org/c/coreboot/+/44343/1/src/mainboard/google/zork/v... PS1, Line 205: (board_sku == 0x5A80000A) || (board_sku == 0x5A80000D) || (board_sku == 0x5A800001)
removed, done. […]
Done
Aaron Durbin has submitted this change. ( https://review.coreboot.org/c/coreboot/+/44343 )
Change subject: mb/google/zork: config ddi for dirinboz ......................................................................
mb/google/zork: config ddi for dirinboz
dirinboz does not support native HDMI, config DDI as below: DDI0: eDP DDI1: DP DDI2: DP
BUG=b:161579679 BRANCH=master TEST=1. emerge-zork coreboot chromeos-bootimage 2. power on proto board successfully
Change-Id: I9dffdf5654680e3c2c0b259ee82a471f8ff14f56 Signed-off-by: Kevin Chiu kevin.chiu@quantatw.com Reviewed-on: https://review.coreboot.org/c/coreboot/+/44343 Tested-by: build bot (Jenkins) no-reply@coreboot.org Reviewed-by: Aaron Durbin adurbin@chromium.org --- M src/mainboard/google/zork/variants/dirinboz/Makefile.inc A src/mainboard/google/zork/variants/dirinboz/variant.c 2 files changed, 45 insertions(+), 0 deletions(-)
Approvals: build bot (Jenkins): Verified Aaron Durbin: Looks good to me, approved
diff --git a/src/mainboard/google/zork/variants/dirinboz/Makefile.inc b/src/mainboard/google/zork/variants/dirinboz/Makefile.inc index 57e7136..51d19fe 100644 --- a/src/mainboard/google/zork/variants/dirinboz/Makefile.inc +++ b/src/mainboard/google/zork/variants/dirinboz/Makefile.inc @@ -3,3 +3,4 @@ subdirs-y += ./spd
ramstage-y += gpio.c +ramstage-y += variant.c diff --git a/src/mainboard/google/zork/variants/dirinboz/variant.c b/src/mainboard/google/zork/variants/dirinboz/variant.c new file mode 100644 index 0000000..ff57f50 --- /dev/null +++ b/src/mainboard/google/zork/variants/dirinboz/variant.c @@ -0,0 +1,44 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +#include <baseboard/variants.h> +#include <console/console.h> +#include <device/device.h> +#include <drivers/i2c/generic/chip.h> +#include <soc/pci_devs.h> +#include <ec/google/chromeec/ec.h> +#include <ec/google/chromeec/i2c_tunnel/chip.h> +#include <string.h> + +/* FIXME: Comments seem to suggest these are not entirely correct. */ +static const fsp_ddi_descriptor non_hdmi_ddi_descriptors[] = { + { + // DDI0, DP0, eDP + .connector_type = EDP, + .aux_index = AUX1, + .hdp_index = HDP1 + }, + { + // DDI1, DP1, DB OPT2 USB-C1 / DB OPT3 MST hub + .connector_type = DP, + .aux_index = AUX2, + .hdp_index = HDP2 + }, + { + // DP2 pins not connected on Dali + // DDI2, DP3, USB-C0 + .connector_type = DP, + .aux_index = AUX4, + .hdp_index = HDP4, + } +}; + +void variant_get_dxio_ddi_descriptors(const fsp_dxio_descriptor **dxio_descs, + size_t *dxio_num, + const fsp_ddi_descriptor **ddi_descs, + size_t *ddi_num) +{ + + *dxio_descs = baseboard_get_dxio_descriptors(dxio_num); + *ddi_descs = &non_hdmi_ddi_descriptors[0]; + *ddi_num = ARRAY_SIZE(non_hdmi_ddi_descriptors); +}