7 CB:31250 ("soc/intel/cannonlake: Configure GPIOs again after FSP-S is
8 done") introduced a workaround in coreboot for `soc/intel/cannonlake`
9 platforms to save and restore GPIO configuration performed by
10 mainboard across call to FSP Silicon Init (FSP-S). This workaround was
11 required because FSP-S was configuring GPIOs differently than
12 mainboard resulting in boot and runtime issues because of
13 misconfigured GPIOs. This issue was observed on `google/hatch`
14 mainboard and was raised with Intel to get the FSP behavior
15 fixed. Until the fix in FSP was available, this workaround was used to
16 ensure that the mainboards can operate correctly and were not impacted
17 by the GPIO misconfiguration in FSP-S.
19 The issues observed on `google/hatch` mainboard were fixed by adding
20 (if required) and initializing appropriate FSP UPDs. This UPD
21 initialization ensured that FSP did not configure any GPIOs
22 differently than the mainboard configuration. Fixes included:
23 * CB:31375 ("soc/intel/cannonlake: Configure serial debug uart")
24 * CB:31520 ("soc/intel/cannonlake: Assign FSP UPDs for HPD and Data/CLK of DDI ports")
25 * CB:32176 ("mb/google/hatch: Update GPIO settings for SD card and SPI1 Chip select")
26 * CB:34900 ("soc/intel/cnl: Add provision to configure SD controller write protect pin")
28 With the above changes merged, it was verified on `google/hatch`
29 mainboard that the workaround for GPIO reconfiguration was not
30 needed. However, at the time, we missed dropping the workaround in
31 'soc/intel/cannonlake`. Currently, this workaround is used by the
38 As verified on `google/hatch`, FSP v1263 included all UPD additions
39 that were required for addressing this issue.
43 * The workaround can be safely dropped from `soc/intel/cannonlake`
44 only after the above mainboards have verified that FSP-S does not
45 configure any pads differently than the mainboard in coreboot. Since
46 the fix included initialization of FSP UPDs correctly, the above
47 mainboards can use the following diff to check what pads change
51 diff --git a/src/soc/intel/common/block/gpio/gpio.c b/src/soc/intel/common/block/gpio/gpio.c
52 index 28e78fb366..0cce41b316 100644
53 --- a/src/soc/intel/common/block/gpio/gpio.c
54 +++ b/src/soc/intel/common/block/gpio/gpio.c
55 @@ -303,10 +303,10 @@ static void gpio_configure_pad(const struct pad_config *cfg)
56 /* Patch GPIO settings for SoC specifically */
57 soc_pad_conf = soc_gpio_pad_config_fixup(cfg, i, soc_pad_conf);
59 - if (CONFIG(DEBUG_GPIO))
60 + if (soc_pad_conf != pad_conf)
62 - "gpio_padcfg [0x%02x, %02zd] DW%d [0x%08x : 0x%08x"
64 + "%d: gpio_padcfg [0x%02x, %02zd] DW%d [0x%08x : 0x%08x"
65 + " : 0x%08x]\n", cfg->pad,
66 comm->port, relative_pad_in_comm(comm, cfg->pad), i,
67 pad_conf,/* old value */
68 cfg->pad_config[i],/* value passed from gpio table */
71 Depending upon the pads that are misconfigured by FSP-S, these
72 mainboards will have to set UPDs appropriately. Once this is verified
73 by the above mainboards, the workaround implemented in CB:31250 can be
76 * The fix implemented in FSP/coreboot for `soc/intel/cannonlake`
77 platforms is not really the right long term solution for the
78 problem. Ideally, FSP should not be touching any GPIO configuration
79 and letting coreboot configure the pads as per mainboard
80 design. This recommendation was accepted and implemented by Intel
81 starting with Jasper Lake and Tiger Lake platforms using a single
82 UPD `GpioOverride` that coreboot can set so that FSP does not change
83 any GPIO configuration. However, this implementation is not
84 backported to any older platforms. Given the issues that we have
85 observed across different platforms, the second proposal is to:
87 - Add a Kconfig `CHECK_GPIO_CONFIG_CHANGES` that enables checks
88 in coreboot to stash GPIO pad configuration before various calls
89 to FSP and compares the configuration on return from FSP.
90 - This will have to be implemented as part of
91 drivers/intel/fsp/fsp2_0/ to check for the above config selection
92 and make callbacks `gpio_snapshot()` and `gpio_verify_snapshot()`
93 to identify and print information about pads that have changed
94 configuration after calls to FSP.
95 - This config can be kept disabled by default and mainboard
96 developers can enable them as and when required for debug.
97 - This will be helpful not just for the `soc/intel/cannonlake`
98 platforms that want to get rid of the above workaround, but also
99 for all future platforms using FSP to identify and catch any GPIO
100 misconfigurations that might slip in to any platforms (in case the
101 `GpioOverride` UPD is not honored by any code path within FSP).