* [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config @ 2022-03-14 9:52 Yann Gautier 2022-03-14 11:17 ` Ulf Hansson 2022-03-14 12:55 ` [PATCH v2] " Yann Gautier 0 siblings, 2 replies; 14+ messages in thread From: Yann Gautier @ 2022-03-14 9:52 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek, Yann Gautier Add a disable_keep_power field in variant_data struct. The MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. It is only set to true for stm32_sdmmc variants. The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. It doesn't correctly support low power, and we need to unbind the wifi driver before a suspend sequence. But the wifi chip firmware is then lost, and the communication with SDIO fails if MMC_PM_KEEP_POWER is enabled. The flag can still be enabled through DT property: keep-power-in-suspend. Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> --- Resend the patch alone. It was previoulsy in a series [1] for which the other patches will be reworked. [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ drivers/mmc/host/mmci.c | 5 ++++- drivers/mmc/host/mmci.h | 1 + 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 45b8608c935c..0e2f2f5d6a52 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { .busy_detect = true, .busy_detect_flag = MCI_STM32_BUSYD0, .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, + .disable_keep_power = true, .init = sdmmc_variant_init, }; @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { .busy_detect = true, .busy_detect_flag = MCI_STM32_BUSYD0, .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, + .disable_keep_power = true, .init = sdmmc_variant_init, }; @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; /* We support these PM capabilities. */ - mmc->pm_caps |= MMC_PM_KEEP_POWER; + if (!variant->disable_keep_power) + mmc->pm_caps |= MMC_PM_KEEP_POWER; /* * We can do SGIO diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h index e1a9b96a3396..2cad1ef9766a 100644 --- a/drivers/mmc/host/mmci.h +++ b/drivers/mmc/host/mmci.h @@ -361,6 +361,7 @@ struct variant_data { u32 opendrain; u8 dma_lli:1; u32 stm32_idmabsize_mask; + u8 disable_keep_power:1; void (*init)(struct mmci_host *host); }; -- 2.25.1 ^ permalink raw reply related [flat|nested] 14+ messages in thread
* Re: [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 9:52 [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config Yann Gautier @ 2022-03-14 11:17 ` Ulf Hansson 2022-03-14 12:20 ` Yann Gautier 2022-03-14 12:55 ` [PATCH v2] " Yann Gautier 1 sibling, 1 reply; 14+ messages in thread From: Ulf Hansson @ 2022-03-14 11:17 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On Mon, 14 Mar 2022 at 10:53, Yann Gautier <yann.gautier@foss.st.com> wrote: > > Add a disable_keep_power field in variant_data struct. The > MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > It is only set to true for stm32_sdmmc variants. > > The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > It doesn't correctly support low power, and we need to unbind the wifi > driver before a suspend sequence. But the wifi chip firmware is then > lost, and the communication with SDIO fails if MMC_PM_KEEP_POWER is > enabled. So the platform supports to maintain the power for the embedded wifi chip during system suspend, but the SDIO func driver (for the WiFi chip) doesn't implement its part correctly. Did I get that right? In that case, it sounds to me like we should try to fix the support for power management in the SDIO func driver instead, no? I am happy to help with guidance/review if that is needed. What SDIO func driver is this about? Kind regards Uffe > The flag can still be enabled through DT property: keep-power-in-suspend. > > Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> > --- > Resend the patch alone. It was previoulsy in a series [1] for which the > other patches will be reworked. > > [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ > > drivers/mmc/host/mmci.c | 5 ++++- > drivers/mmc/host/mmci.h | 1 + > 2 files changed, 5 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c > index 45b8608c935c..0e2f2f5d6a52 100644 > --- a/drivers/mmc/host/mmci.c > +++ b/drivers/mmc/host/mmci.c > @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, > host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; > > /* We support these PM capabilities. */ > - mmc->pm_caps |= MMC_PM_KEEP_POWER; > + if (!variant->disable_keep_power) > + mmc->pm_caps |= MMC_PM_KEEP_POWER; > > /* > * We can do SGIO > diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h > index e1a9b96a3396..2cad1ef9766a 100644 > --- a/drivers/mmc/host/mmci.h > +++ b/drivers/mmc/host/mmci.h > @@ -361,6 +361,7 @@ struct variant_data { > u32 opendrain; > u8 dma_lli:1; > u32 stm32_idmabsize_mask; > + u8 disable_keep_power:1; > void (*init)(struct mmci_host *host); > }; > > -- > 2.25.1 > ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 11:17 ` Ulf Hansson @ 2022-03-14 12:20 ` Yann Gautier 2022-03-14 12:53 ` Ulf Hansson 0 siblings, 1 reply; 14+ messages in thread From: Yann Gautier @ 2022-03-14 12:20 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On 3/14/22 12:17, Ulf Hansson wrote: > On Mon, 14 Mar 2022 at 10:53, Yann Gautier <yann.gautier@foss.st.com> wrote: >> >> Add a disable_keep_power field in variant_data struct. The >> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. >> It is only set to true for stm32_sdmmc variants. >> >> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. >> It doesn't correctly support low power, and we need to unbind the wifi >> driver before a suspend sequence. But the wifi chip firmware is then >> lost, and the communication with SDIO fails if MMC_PM_KEEP_POWER is >> enabled. > > So the platform supports to maintain the power for the embedded wifi > chip during system suspend, but the SDIO func driver (for the WiFi > chip) doesn't implement its part correctly. Did I get that right? > > In that case, it sounds to me like we should try to fix the support > for power management in the SDIO func driver instead, no? > I am happy to help with guidance/review if that is needed. What SDIO > func driver is this about? > > Kind regards > Uffe > Hi Ulf, I blindly pushed the patch without rechecking it. I rephrased it in our downstream to better explain the issue: The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. It doesn't correctly support low power on this board. The Wifi chip awaits an always-on regulator, but it was connected to v3v3 which is off in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. If it's OK for you, I'll resend the patch with the updated commit message. Best regards, Yann >> The flag can still be enabled through DT property: keep-power-in-suspend. >> >> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> >> --- >> Resend the patch alone. It was previoulsy in a series [1] for which the >> other patches will be reworked. >> >> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ >> >> drivers/mmc/host/mmci.c | 5 ++++- >> drivers/mmc/host/mmci.h | 1 + >> 2 files changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c >> index 45b8608c935c..0e2f2f5d6a52 100644 >> --- a/drivers/mmc/host/mmci.c >> +++ b/drivers/mmc/host/mmci.c >> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, >> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; >> >> /* We support these PM capabilities. */ >> - mmc->pm_caps |= MMC_PM_KEEP_POWER; >> + if (!variant->disable_keep_power) >> + mmc->pm_caps |= MMC_PM_KEEP_POWER; >> >> /* >> * We can do SGIO >> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h >> index e1a9b96a3396..2cad1ef9766a 100644 >> --- a/drivers/mmc/host/mmci.h >> +++ b/drivers/mmc/host/mmci.h >> @@ -361,6 +361,7 @@ struct variant_data { >> u32 opendrain; >> u8 dma_lli:1; >> u32 stm32_idmabsize_mask; >> + u8 disable_keep_power:1; >> void (*init)(struct mmci_host *host); >> }; >> >> -- >> 2.25.1 >> ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 12:20 ` Yann Gautier @ 2022-03-14 12:53 ` Ulf Hansson 0 siblings, 0 replies; 14+ messages in thread From: Ulf Hansson @ 2022-03-14 12:53 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On Mon, 14 Mar 2022 at 13:20, Yann Gautier <yann.gautier@foss.st.com> wrote: > > On 3/14/22 12:17, Ulf Hansson wrote: > > On Mon, 14 Mar 2022 at 10:53, Yann Gautier <yann.gautier@foss.st.com> wrote: > >> > >> Add a disable_keep_power field in variant_data struct. The > >> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > >> It is only set to true for stm32_sdmmc variants. > >> > >> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > >> It doesn't correctly support low power, and we need to unbind the wifi > >> driver before a suspend sequence. But the wifi chip firmware is then > >> lost, and the communication with SDIO fails if MMC_PM_KEEP_POWER is > >> enabled. > > > > So the platform supports to maintain the power for the embedded wifi > > chip during system suspend, but the SDIO func driver (for the WiFi > > chip) doesn't implement its part correctly. Did I get that right? > > > > In that case, it sounds to me like we should try to fix the support > > for power management in the SDIO func driver instead, no? > > I am happy to help with guidance/review if that is needed. What SDIO > > func driver is this about? > > > > Kind regards > > Uffe > > > > Hi Ulf, > > I blindly pushed the patch without rechecking it. > I rephrased it in our downstream to better explain the issue: > > The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > It doesn't correctly support low power on this board. The Wifi chip > awaits an always-on regulator, but it was connected to v3v3 which is off > in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > > If it's OK for you, I'll resend the patch with the updated commit message. Yes, that works fine for me! [...] Kind regards Uffe ^ permalink raw reply [flat|nested] 14+ messages in thread
* [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 9:52 [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config Yann Gautier 2022-03-14 11:17 ` Ulf Hansson @ 2022-03-14 12:55 ` Yann Gautier 2022-03-14 13:03 ` Ulf Hansson 2022-03-14 13:05 ` Philipp Zabel 1 sibling, 2 replies; 14+ messages in thread From: Yann Gautier @ 2022-03-14 12:55 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek, Yann Gautier Add a disable_keep_power field in variant_data struct. The MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. It is only set to true for stm32_sdmmc variants. The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. It doesn't correctly support low power on this board. The Wifi chip awaits an always-on regulator, but it was connected to v3v3 which is off in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. The flag can still be enabled through DT property: keep-power-in-suspend. Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> --- Update in v2: Reword commit message to better explain the issue. Resend the patch alone. It was previoulsy in a series [1] for which the other patches will be reworked. [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ drivers/mmc/host/mmci.c | 5 ++++- drivers/mmc/host/mmci.h | 1 + 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 45b8608c935c..0e2f2f5d6a52 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { .busy_detect = true, .busy_detect_flag = MCI_STM32_BUSYD0, .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, + .disable_keep_power = true, .init = sdmmc_variant_init, }; @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { .busy_detect = true, .busy_detect_flag = MCI_STM32_BUSYD0, .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, + .disable_keep_power = true, .init = sdmmc_variant_init, }; @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; /* We support these PM capabilities. */ - mmc->pm_caps |= MMC_PM_KEEP_POWER; + if (!variant->disable_keep_power) + mmc->pm_caps |= MMC_PM_KEEP_POWER; /* * We can do SGIO diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h index e1a9b96a3396..2cad1ef9766a 100644 --- a/drivers/mmc/host/mmci.h +++ b/drivers/mmc/host/mmci.h @@ -361,6 +361,7 @@ struct variant_data { u32 opendrain; u8 dma_lli:1; u32 stm32_idmabsize_mask; + u8 disable_keep_power:1; void (*init)(struct mmci_host *host); }; -- 2.25.1 ^ permalink raw reply related [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 12:55 ` [PATCH v2] " Yann Gautier @ 2022-03-14 13:03 ` Ulf Hansson 2022-03-14 14:34 ` Yann Gautier 2022-03-14 13:05 ` Philipp Zabel 1 sibling, 1 reply; 14+ messages in thread From: Ulf Hansson @ 2022-03-14 13:03 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: > > Add a disable_keep_power field in variant_data struct. The > MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > It is only set to true for stm32_sdmmc variants. > > The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > It doesn't correctly support low power on this board. The Wifi chip > awaits an always-on regulator, but it was connected to v3v3 which is off > in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. Just to make sure I get this correct. Why can't the regulator stay on during system suspend? The point is, we don't need an always on regulator to cope with this. Kind regards Uffe > > The flag can still be enabled through DT property: > keep-power-in-suspend. > > Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> > --- > Update in v2: > Reword commit message to better explain the issue. > > Resend the patch alone. It was previoulsy in a series [1] for which the > other patches will be reworked. > > [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ > > drivers/mmc/host/mmci.c | 5 ++++- > drivers/mmc/host/mmci.h | 1 + > 2 files changed, 5 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c > index 45b8608c935c..0e2f2f5d6a52 100644 > --- a/drivers/mmc/host/mmci.c > +++ b/drivers/mmc/host/mmci.c > @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, > host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; > > /* We support these PM capabilities. */ > - mmc->pm_caps |= MMC_PM_KEEP_POWER; > + if (!variant->disable_keep_power) > + mmc->pm_caps |= MMC_PM_KEEP_POWER; > > /* > * We can do SGIO > diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h > index e1a9b96a3396..2cad1ef9766a 100644 > --- a/drivers/mmc/host/mmci.h > +++ b/drivers/mmc/host/mmci.h > @@ -361,6 +361,7 @@ struct variant_data { > u32 opendrain; > u8 dma_lli:1; > u32 stm32_idmabsize_mask; > + u8 disable_keep_power:1; > void (*init)(struct mmci_host *host); > }; > > -- > 2.25.1 > ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 13:03 ` Ulf Hansson @ 2022-03-14 14:34 ` Yann Gautier 2022-03-14 16:18 ` Ulf Hansson 0 siblings, 1 reply; 14+ messages in thread From: Yann Gautier @ 2022-03-14 14:34 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On 3/14/22 14:03, Ulf Hansson wrote: > On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: >> >> Add a disable_keep_power field in variant_data struct. The >> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. >> It is only set to true for stm32_sdmmc variants. >> >> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. >> It doesn't correctly support low power on this board. The Wifi chip >> awaits an always-on regulator, but it was connected to v3v3 which is off >> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > > Just to make sure I get this correct. > > Why can't the regulator stay on during system suspend? The point is, > we don't need an always on regulator to cope with this. > > Kind regards > Uffe Hi Ulf, This v3v3 regulator powers most of the devices on this board. So we need to switch it off to gain power in suspend mode. Yann > >> >> The flag can still be enabled through DT property: >> keep-power-in-suspend. >> >> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> >> --- >> Update in v2: >> Reword commit message to better explain the issue. >> >> Resend the patch alone. It was previoulsy in a series [1] for which the >> other patches will be reworked. >> >> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ >> >> drivers/mmc/host/mmci.c | 5 ++++- >> drivers/mmc/host/mmci.h | 1 + >> 2 files changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c >> index 45b8608c935c..0e2f2f5d6a52 100644 >> --- a/drivers/mmc/host/mmci.c >> +++ b/drivers/mmc/host/mmci.c >> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, >> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; >> >> /* We support these PM capabilities. */ >> - mmc->pm_caps |= MMC_PM_KEEP_POWER; >> + if (!variant->disable_keep_power) >> + mmc->pm_caps |= MMC_PM_KEEP_POWER; >> >> /* >> * We can do SGIO >> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h >> index e1a9b96a3396..2cad1ef9766a 100644 >> --- a/drivers/mmc/host/mmci.h >> +++ b/drivers/mmc/host/mmci.h >> @@ -361,6 +361,7 @@ struct variant_data { >> u32 opendrain; >> u8 dma_lli:1; >> u32 stm32_idmabsize_mask; >> + u8 disable_keep_power:1; >> void (*init)(struct mmci_host *host); >> }; >> >> -- >> 2.25.1 >> ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 14:34 ` Yann Gautier @ 2022-03-14 16:18 ` Ulf Hansson 2022-03-14 16:51 ` Yann Gautier 0 siblings, 1 reply; 14+ messages in thread From: Ulf Hansson @ 2022-03-14 16:18 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On Mon, 14 Mar 2022 at 15:34, Yann Gautier <yann.gautier@foss.st.com> wrote: > > On 3/14/22 14:03, Ulf Hansson wrote: > > On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: > >> > >> Add a disable_keep_power field in variant_data struct. The > >> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > >> It is only set to true for stm32_sdmmc variants. > >> > >> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > >> It doesn't correctly support low power on this board. The Wifi chip > >> awaits an always-on regulator, but it was connected to v3v3 which is off > >> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > > > > Just to make sure I get this correct. > > > > Why can't the regulator stay on during system suspend? The point is, > > we don't need an always on regulator to cope with this. > > > > Kind regards > > Uffe > > Hi Ulf, > > This v3v3 regulator powers most of the devices on this board. So we need > to switch it off to gain power in suspend mode. I see. Thanks for sharing that information. The MMC_PM_KEEP_POWER flag is there to describe what is supported by the platform/host. It doesn't mean that the card *must* stay powered on during system suspend. Instead that depends on whether system wakeup for the SDIO/WiFi is supported too - and if that is enabled by userspace. If not, the regulator will be turned off for the SDIO card during system suspend. Assuming the regulator is implemented as a proper regulator and can remain on during system suspend, the right thing would be to keep the MMC_PM_KEEP_POWER flag around. Kind regards Uffe > > > Yann > > > > >> > >> The flag can still be enabled through DT property: > >> keep-power-in-suspend. > >> > >> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> > >> --- > >> Update in v2: > >> Reword commit message to better explain the issue. > >> > >> Resend the patch alone. It was previoulsy in a series [1] for which the > >> other patches will be reworked. > >> > >> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ > >> > >> drivers/mmc/host/mmci.c | 5 ++++- > >> drivers/mmc/host/mmci.h | 1 + > >> 2 files changed, 5 insertions(+), 1 deletion(-) > >> > >> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c > >> index 45b8608c935c..0e2f2f5d6a52 100644 > >> --- a/drivers/mmc/host/mmci.c > >> +++ b/drivers/mmc/host/mmci.c > >> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { > >> .busy_detect = true, > >> .busy_detect_flag = MCI_STM32_BUSYD0, > >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > >> + .disable_keep_power = true, > >> .init = sdmmc_variant_init, > >> }; > >> > >> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { > >> .busy_detect = true, > >> .busy_detect_flag = MCI_STM32_BUSYD0, > >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > >> + .disable_keep_power = true, > >> .init = sdmmc_variant_init, > >> }; > >> > >> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, > >> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; > >> > >> /* We support these PM capabilities. */ > >> - mmc->pm_caps |= MMC_PM_KEEP_POWER; > >> + if (!variant->disable_keep_power) > >> + mmc->pm_caps |= MMC_PM_KEEP_POWER; > >> > >> /* > >> * We can do SGIO > >> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h > >> index e1a9b96a3396..2cad1ef9766a 100644 > >> --- a/drivers/mmc/host/mmci.h > >> +++ b/drivers/mmc/host/mmci.h > >> @@ -361,6 +361,7 @@ struct variant_data { > >> u32 opendrain; > >> u8 dma_lli:1; > >> u32 stm32_idmabsize_mask; > >> + u8 disable_keep_power:1; > >> void (*init)(struct mmci_host *host); > >> }; > >> > >> -- > >> 2.25.1 > >> > ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 16:18 ` Ulf Hansson @ 2022-03-14 16:51 ` Yann Gautier 2022-03-15 9:07 ` Ulf Hansson 0 siblings, 1 reply; 14+ messages in thread From: Yann Gautier @ 2022-03-14 16:51 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On 3/14/22 17:18, Ulf Hansson wrote: > On Mon, 14 Mar 2022 at 15:34, Yann Gautier <yann.gautier@foss.st.com> wrote: >> >> On 3/14/22 14:03, Ulf Hansson wrote: >>> On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: >>>> >>>> Add a disable_keep_power field in variant_data struct. The >>>> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. >>>> It is only set to true for stm32_sdmmc variants. >>>> >>>> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. >>>> It doesn't correctly support low power on this board. The Wifi chip >>>> awaits an always-on regulator, but it was connected to v3v3 which is off >>>> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. >>> >>> Just to make sure I get this correct. >>> >>> Why can't the regulator stay on during system suspend? The point is, >>> we don't need an always on regulator to cope with this. >>> >>> Kind regards >>> Uffe >> >> Hi Ulf, >> >> This v3v3 regulator powers most of the devices on this board. So we need >> to switch it off to gain power in suspend mode. > > I see. Thanks for sharing that information. > > The MMC_PM_KEEP_POWER flag is there to describe what is supported by > the platform/host. It doesn't mean that the card *must* stay powered > on during system suspend. Instead that depends on whether system > wakeup for the SDIO/WiFi is supported too - and if that is enabled by > userspace. If not, the regulator will be turned off for the SDIO card > during system suspend. > > Assuming the regulator is implemented as a proper regulator and can > remain on during system suspend, the right thing would be to keep the > MMC_PM_KEEP_POWER flag around. > > Kind regards > Uffe > OK, but in the wifi driver we use on this platform (brcmfmac), the suspend/resume functions (brcmf_ops_sdio_suspend/brcmf_ops_sdio_resume) use the flag to check regu was off, and then call probe function during resume, to re-init Wifi chip and reload its firmware. Best regards, Yann >> >> >> Yann >> >>> >>>> >>>> The flag can still be enabled through DT property: >>>> keep-power-in-suspend. >>>> >>>> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> >>>> --- >>>> Update in v2: >>>> Reword commit message to better explain the issue. >>>> >>>> Resend the patch alone. It was previoulsy in a series [1] for which the >>>> other patches will be reworked. >>>> >>>> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ >>>> >>>> drivers/mmc/host/mmci.c | 5 ++++- >>>> drivers/mmc/host/mmci.h | 1 + >>>> 2 files changed, 5 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c >>>> index 45b8608c935c..0e2f2f5d6a52 100644 >>>> --- a/drivers/mmc/host/mmci.c >>>> +++ b/drivers/mmc/host/mmci.c >>>> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { >>>> .busy_detect = true, >>>> .busy_detect_flag = MCI_STM32_BUSYD0, >>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >>>> + .disable_keep_power = true, >>>> .init = sdmmc_variant_init, >>>> }; >>>> >>>> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { >>>> .busy_detect = true, >>>> .busy_detect_flag = MCI_STM32_BUSYD0, >>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >>>> + .disable_keep_power = true, >>>> .init = sdmmc_variant_init, >>>> }; >>>> >>>> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, >>>> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; >>>> >>>> /* We support these PM capabilities. */ >>>> - mmc->pm_caps |= MMC_PM_KEEP_POWER; >>>> + if (!variant->disable_keep_power) >>>> + mmc->pm_caps |= MMC_PM_KEEP_POWER; >>>> >>>> /* >>>> * We can do SGIO >>>> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h >>>> index e1a9b96a3396..2cad1ef9766a 100644 >>>> --- a/drivers/mmc/host/mmci.h >>>> +++ b/drivers/mmc/host/mmci.h >>>> @@ -361,6 +361,7 @@ struct variant_data { >>>> u32 opendrain; >>>> u8 dma_lli:1; >>>> u32 stm32_idmabsize_mask; >>>> + u8 disable_keep_power:1; >>>> void (*init)(struct mmci_host *host); >>>> }; >>>> >>>> -- >>>> 2.25.1 >>>> >> ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 16:51 ` Yann Gautier @ 2022-03-15 9:07 ` Ulf Hansson 2022-03-15 10:07 ` Yann Gautier 0 siblings, 1 reply; 14+ messages in thread From: Ulf Hansson @ 2022-03-15 9:07 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On Mon, 14 Mar 2022 at 17:52, Yann Gautier <yann.gautier@foss.st.com> wrote: > > On 3/14/22 17:18, Ulf Hansson wrote: > > On Mon, 14 Mar 2022 at 15:34, Yann Gautier <yann.gautier@foss.st.com> wrote: > >> > >> On 3/14/22 14:03, Ulf Hansson wrote: > >>> On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: > >>>> > >>>> Add a disable_keep_power field in variant_data struct. The > >>>> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > >>>> It is only set to true for stm32_sdmmc variants. > >>>> > >>>> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > >>>> It doesn't correctly support low power on this board. The Wifi chip > >>>> awaits an always-on regulator, but it was connected to v3v3 which is off > >>>> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > >>> > >>> Just to make sure I get this correct. > >>> > >>> Why can't the regulator stay on during system suspend? The point is, > >>> we don't need an always on regulator to cope with this. > >>> > >>> Kind regards > >>> Uffe > >> > >> Hi Ulf, > >> > >> This v3v3 regulator powers most of the devices on this board. So we need > >> to switch it off to gain power in suspend mode. > > > > I see. Thanks for sharing that information. > > > > The MMC_PM_KEEP_POWER flag is there to describe what is supported by > > the platform/host. It doesn't mean that the card *must* stay powered > > on during system suspend. Instead that depends on whether system > > wakeup for the SDIO/WiFi is supported too - and if that is enabled by > > userspace. If not, the regulator will be turned off for the SDIO card > > during system suspend. > > > > Assuming the regulator is implemented as a proper regulator and can > > remain on during system suspend, the right thing would be to keep the > > MMC_PM_KEEP_POWER flag around. > > > > Kind regards > > Uffe > > > > OK, but in the wifi driver we use on this platform (brcmfmac), the > suspend/resume functions (brcmf_ops_sdio_suspend/brcmf_ops_sdio_resume) > use the flag to check regu was off, and then call probe function during > resume, to re-init Wifi chip and reload its firmware. I had a closer look at the brcmfmac driver, thanks for the pointers. In my opinion, I think we should change the brcmfmac driver, so it decides to power off the SDIO card, unless the WiFi chip is configured to serve us with system wakeups. I can send a patch for brcmfmac that we can try, unless you want to send it yourself? > > > Best regards, > Yann Kind regards Uffe > > >> > >> > >> Yann > >> > >>> > >>>> > >>>> The flag can still be enabled through DT property: > >>>> keep-power-in-suspend. > >>>> > >>>> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> > >>>> --- > >>>> Update in v2: > >>>> Reword commit message to better explain the issue. > >>>> > >>>> Resend the patch alone. It was previoulsy in a series [1] for which the > >>>> other patches will be reworked. > >>>> > >>>> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ > >>>> > >>>> drivers/mmc/host/mmci.c | 5 ++++- > >>>> drivers/mmc/host/mmci.h | 1 + > >>>> 2 files changed, 5 insertions(+), 1 deletion(-) > >>>> > >>>> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c > >>>> index 45b8608c935c..0e2f2f5d6a52 100644 > >>>> --- a/drivers/mmc/host/mmci.c > >>>> +++ b/drivers/mmc/host/mmci.c > >>>> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { > >>>> .busy_detect = true, > >>>> .busy_detect_flag = MCI_STM32_BUSYD0, > >>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > >>>> + .disable_keep_power = true, > >>>> .init = sdmmc_variant_init, > >>>> }; > >>>> > >>>> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { > >>>> .busy_detect = true, > >>>> .busy_detect_flag = MCI_STM32_BUSYD0, > >>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > >>>> + .disable_keep_power = true, > >>>> .init = sdmmc_variant_init, > >>>> }; > >>>> > >>>> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, > >>>> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; > >>>> > >>>> /* We support these PM capabilities. */ > >>>> - mmc->pm_caps |= MMC_PM_KEEP_POWER; > >>>> + if (!variant->disable_keep_power) > >>>> + mmc->pm_caps |= MMC_PM_KEEP_POWER; > >>>> > >>>> /* > >>>> * We can do SGIO > >>>> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h > >>>> index e1a9b96a3396..2cad1ef9766a 100644 > >>>> --- a/drivers/mmc/host/mmci.h > >>>> +++ b/drivers/mmc/host/mmci.h > >>>> @@ -361,6 +361,7 @@ struct variant_data { > >>>> u32 opendrain; > >>>> u8 dma_lli:1; > >>>> u32 stm32_idmabsize_mask; > >>>> + u8 disable_keep_power:1; > >>>> void (*init)(struct mmci_host *host); > >>>> }; > >>>> > >>>> -- > >>>> 2.25.1 > >>>> > >> > ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-15 9:07 ` Ulf Hansson @ 2022-03-15 10:07 ` Yann Gautier 2022-03-15 12:30 ` Ulf Hansson 0 siblings, 1 reply; 14+ messages in thread From: Yann Gautier @ 2022-03-15 10:07 UTC (permalink / raw) To: Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek, Christophe ROULLIER-SCND-02 On 3/15/22 10:07, Ulf Hansson wrote: > On Mon, 14 Mar 2022 at 17:52, Yann Gautier <yann.gautier@foss.st.com> wrote: >> >> On 3/14/22 17:18, Ulf Hansson wrote: >>> On Mon, 14 Mar 2022 at 15:34, Yann Gautier <yann.gautier@foss.st.com> wrote: >>>> >>>> On 3/14/22 14:03, Ulf Hansson wrote: >>>>> On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: >>>>>> >>>>>> Add a disable_keep_power field in variant_data struct. The >>>>>> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. >>>>>> It is only set to true for stm32_sdmmc variants. >>>>>> >>>>>> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. >>>>>> It doesn't correctly support low power on this board. The Wifi chip >>>>>> awaits an always-on regulator, but it was connected to v3v3 which is off >>>>>> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. >>>>> >>>>> Just to make sure I get this correct. >>>>> >>>>> Why can't the regulator stay on during system suspend? The point is, >>>>> we don't need an always on regulator to cope with this. >>>>> >>>>> Kind regards >>>>> Uffe >>>> >>>> Hi Ulf, >>>> >>>> This v3v3 regulator powers most of the devices on this board. So we need >>>> to switch it off to gain power in suspend mode. >>> >>> I see. Thanks for sharing that information. >>> >>> The MMC_PM_KEEP_POWER flag is there to describe what is supported by >>> the platform/host. It doesn't mean that the card *must* stay powered >>> on during system suspend. Instead that depends on whether system >>> wakeup for the SDIO/WiFi is supported too - and if that is enabled by >>> userspace. If not, the regulator will be turned off for the SDIO card >>> during system suspend. >>> >>> Assuming the regulator is implemented as a proper regulator and can >>> remain on during system suspend, the right thing would be to keep the >>> MMC_PM_KEEP_POWER flag around. >>> >>> Kind regards >>> Uffe >>> >> >> OK, but in the wifi driver we use on this platform (brcmfmac), the >> suspend/resume functions (brcmf_ops_sdio_suspend/brcmf_ops_sdio_resume) >> use the flag to check regu was off, and then call probe function during >> resume, to re-init Wifi chip and reload its firmware. > > I had a closer look at the brcmfmac driver, thanks for the pointers. > > In my opinion, I think we should change the brcmfmac driver, so it > decides to power off the SDIO card, unless the WiFi chip is configured > to serve us with system wakeups. > > I can send a patch for brcmfmac that we can try, unless you want to > send it yourself? > Hi Ulf, If you already have an idea of the patch in the brcmfmac driver, can you propose something? We'll be able to test it at our side. Thanks, Yann >> >> >> Best regards, >> Yann > > Kind regards > Uffe > >> >>>> >>>> >>>> Yann >>>> >>>>> >>>>>> >>>>>> The flag can still be enabled through DT property: >>>>>> keep-power-in-suspend. >>>>>> >>>>>> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> >>>>>> --- >>>>>> Update in v2: >>>>>> Reword commit message to better explain the issue. >>>>>> >>>>>> Resend the patch alone. It was previoulsy in a series [1] for which the >>>>>> other patches will be reworked. >>>>>> >>>>>> [1] https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ >>>>>> >>>>>> drivers/mmc/host/mmci.c | 5 ++++- >>>>>> drivers/mmc/host/mmci.h | 1 + >>>>>> 2 files changed, 5 insertions(+), 1 deletion(-) >>>>>> >>>>>> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c >>>>>> index 45b8608c935c..0e2f2f5d6a52 100644 >>>>>> --- a/drivers/mmc/host/mmci.c >>>>>> +++ b/drivers/mmc/host/mmci.c >>>>>> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = { >>>>>> .busy_detect = true, >>>>>> .busy_detect_flag = MCI_STM32_BUSYD0, >>>>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >>>>>> + .disable_keep_power = true, >>>>>> .init = sdmmc_variant_init, >>>>>> }; >>>>>> >>>>>> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 = { >>>>>> .busy_detect = true, >>>>>> .busy_detect_flag = MCI_STM32_BUSYD0, >>>>>> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >>>>>> + .disable_keep_power = true, >>>>>> .init = sdmmc_variant_init, >>>>>> }; >>>>>> >>>>>> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, >>>>>> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; >>>>>> >>>>>> /* We support these PM capabilities. */ >>>>>> - mmc->pm_caps |= MMC_PM_KEEP_POWER; >>>>>> + if (!variant->disable_keep_power) >>>>>> + mmc->pm_caps |= MMC_PM_KEEP_POWER; >>>>>> >>>>>> /* >>>>>> * We can do SGIO >>>>>> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h >>>>>> index e1a9b96a3396..2cad1ef9766a 100644 >>>>>> --- a/drivers/mmc/host/mmci.h >>>>>> +++ b/drivers/mmc/host/mmci.h >>>>>> @@ -361,6 +361,7 @@ struct variant_data { >>>>>> u32 opendrain; >>>>>> u8 dma_lli:1; >>>>>> u32 stm32_idmabsize_mask; >>>>>> + u8 disable_keep_power:1; >>>>>> void (*init)(struct mmci_host *host); >>>>>> }; >>>>>> >>>>>> -- >>>>>> 2.25.1 >>>>>> >>>> >> ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-15 10:07 ` Yann Gautier @ 2022-03-15 12:30 ` Ulf Hansson 0 siblings, 0 replies; 14+ messages in thread From: Ulf Hansson @ 2022-03-15 12:30 UTC (permalink / raw) To: Yann Gautier Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Philipp Zabel, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek, Christophe ROULLIER-SCND-02 On Tue, 15 Mar 2022 at 11:08, Yann Gautier <yann.gautier@foss.st.com> wrote: > > On 3/15/22 10:07, Ulf Hansson wrote: > > On Mon, 14 Mar 2022 at 17:52, Yann Gautier <yann.gautier@foss.st.com> wrote: > >> > >> On 3/14/22 17:18, Ulf Hansson wrote: > >>> On Mon, 14 Mar 2022 at 15:34, Yann Gautier <yann.gautier@foss.st.com> wrote: > >>>> > >>>> On 3/14/22 14:03, Ulf Hansson wrote: > >>>>> On Mon, 14 Mar 2022 at 13:56, Yann Gautier <yann.gautier@foss.st.com> wrote: > >>>>>> > >>>>>> Add a disable_keep_power field in variant_data struct. The > >>>>>> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not set. > >>>>>> It is only set to true for stm32_sdmmc variants. > >>>>>> > >>>>>> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi chip. > >>>>>> It doesn't correctly support low power on this board. The Wifi chip > >>>>>> awaits an always-on regulator, but it was connected to v3v3 which is off > >>>>>> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > >>>>> > >>>>> Just to make sure I get this correct. > >>>>> > >>>>> Why can't the regulator stay on during system suspend? The point is, > >>>>> we don't need an always on regulator to cope with this. > >>>>> > >>>>> Kind regards > >>>>> Uffe > >>>> > >>>> Hi Ulf, > >>>> > >>>> This v3v3 regulator powers most of the devices on this board. So we need > >>>> to switch it off to gain power in suspend mode. > >>> > >>> I see. Thanks for sharing that information. > >>> > >>> The MMC_PM_KEEP_POWER flag is there to describe what is supported by > >>> the platform/host. It doesn't mean that the card *must* stay powered > >>> on during system suspend. Instead that depends on whether system > >>> wakeup for the SDIO/WiFi is supported too - and if that is enabled by > >>> userspace. If not, the regulator will be turned off for the SDIO card > >>> during system suspend. > >>> > >>> Assuming the regulator is implemented as a proper regulator and can > >>> remain on during system suspend, the right thing would be to keep the > >>> MMC_PM_KEEP_POWER flag around. > >>> > >>> Kind regards > >>> Uffe > >>> > >> > >> OK, but in the wifi driver we use on this platform (brcmfmac), the > >> suspend/resume functions (brcmf_ops_sdio_suspend/brcmf_ops_sdio_resume) > >> use the flag to check regu was off, and then call probe function during > >> resume, to re-init Wifi chip and reload its firmware. > > > > I had a closer look at the brcmfmac driver, thanks for the pointers. > > > > In my opinion, I think we should change the brcmfmac driver, so it > > decides to power off the SDIO card, unless the WiFi chip is configured > > to serve us with system wakeups. > > > > I can send a patch for brcmfmac that we can try, unless you want to > > send it yourself? > > > > Hi Ulf, > If you already have an idea of the patch in the brcmfmac driver, can you > propose something? Sure, I will post a patch as soon as I can. > We'll be able to test it at our side. Great, I will keep you posted! [...] Kind regards Uffe ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 12:55 ` [PATCH v2] " Yann Gautier 2022-03-14 13:03 ` Ulf Hansson @ 2022-03-14 13:05 ` Philipp Zabel 2022-03-14 13:45 ` Yann Gautier 1 sibling, 1 reply; 14+ messages in thread From: Philipp Zabel @ 2022-03-14 13:05 UTC (permalink / raw) To: Yann Gautier, Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek Hi Yann, On Mo, 2022-03-14 at 13:55 +0100, Yann Gautier wrote: > Add a disable_keep_power field in variant_data struct. The > MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not > set. > It is only set to true for stm32_sdmmc variants. > > The issue was seen on STM32MP157C-DK2 board, which embeds a wifi > chip. > It doesn't correctly support low power on this board. The Wifi chip > awaits an always-on regulator, but it was connected to v3v3 which is > off > in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. > > The flag can still be enabled through DT property: > keep-power-in-suspend. > > Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> > --- > Update in v2: > Reword commit message to better explain the issue. > > Resend the patch alone. It was previoulsy in a series [1] for which > the > other patches will be reworked. > > [1] > https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ > > drivers/mmc/host/mmci.c | 5 ++++- > drivers/mmc/host/mmci.h | 1 + > 2 files changed, 5 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c > index 45b8608c935c..0e2f2f5d6a52 100644 > --- a/drivers/mmc/host/mmci.c > +++ b/drivers/mmc/host/mmci.c > @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = > { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 > = { > .busy_detect = true, > .busy_detect_flag = MCI_STM32_BUSYD0, > .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, > + .disable_keep_power = true, > .init = sdmmc_variant_init, > }; > > @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, > host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; > > /* We support these PM capabilities. */ > - mmc->pm_caps |= MMC_PM_KEEP_POWER; > + if (!variant->disable_keep_power) > + mmc->pm_caps |= MMC_PM_KEEP_POWER; > > /* > * We can do SGIO > diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h > index e1a9b96a3396..2cad1ef9766a 100644 > --- a/drivers/mmc/host/mmci.h > +++ b/drivers/mmc/host/mmci.h > @@ -361,6 +361,7 @@ struct variant_data { > u32 opendrain; > u8 dma_lli:1; > u32 stm32_idmabsize_mask; > + u8 disable_keep_power:1; There are already four separate bitfields in struct variant_data, why not move this up into one of them? regards Philipp ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v2] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config 2022-03-14 13:05 ` Philipp Zabel @ 2022-03-14 13:45 ` Yann Gautier 0 siblings, 0 replies; 14+ messages in thread From: Yann Gautier @ 2022-03-14 13:45 UTC (permalink / raw) To: Philipp Zabel, Ulf Hansson Cc: linux-kernel, linux-mmc, linux-stm32, Christophe Kerello, Linus Walleij, Ludovic Barre, Russell King, Marek Vasut, kernel, Manivannan Sadhasivam, Grzegorz Szymaszek On 3/14/22 14:05, Philipp Zabel wrote: > Hi Yann, > > On Mo, 2022-03-14 at 13:55 +0100, Yann Gautier wrote: >> Add a disable_keep_power field in variant_data struct. The >> MMC_PM_KEEP_POWER flag will be enabled if disable_keep_power is not >> set. >> It is only set to true for stm32_sdmmc variants. >> >> The issue was seen on STM32MP157C-DK2 board, which embeds a wifi >> chip. >> It doesn't correctly support low power on this board. The Wifi chip >> awaits an always-on regulator, but it was connected to v3v3 which is >> off >> in low-power sequence. MMC_PM_KEEP_POWER should then be disabled. >> >> The flag can still be enabled through DT property: >> keep-power-in-suspend. >> >> Signed-off-by: Yann Gautier <yann.gautier@foss.st.com> >> --- >> Update in v2: >> Reword commit message to better explain the issue. >> >> Resend the patch alone. It was previoulsy in a series [1] for which >> the >> other patches will be reworked. >> >> [1] >> https://lore.kernel.org/lkml/20220304135134.47827-1-yann.gautier@foss.st.com/ >> >> drivers/mmc/host/mmci.c | 5 ++++- >> drivers/mmc/host/mmci.h | 1 + >> 2 files changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c >> index 45b8608c935c..0e2f2f5d6a52 100644 >> --- a/drivers/mmc/host/mmci.c >> +++ b/drivers/mmc/host/mmci.c >> @@ -274,6 +274,7 @@ static struct variant_data variant_stm32_sdmmc = >> { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -301,6 +302,7 @@ static struct variant_data variant_stm32_sdmmcv2 >> = { >> .busy_detect = true, >> .busy_detect_flag = MCI_STM32_BUSYD0, >> .busy_detect_mask = MCI_STM32_BUSYD0ENDMASK, >> + .disable_keep_power = true, >> .init = sdmmc_variant_init, >> }; >> >> @@ -2172,7 +2174,8 @@ static int mmci_probe(struct amba_device *dev, >> host->stop_abort.flags = MMC_RSP_R1B | MMC_CMD_AC; >> >> /* We support these PM capabilities. */ >> - mmc->pm_caps |= MMC_PM_KEEP_POWER; >> + if (!variant->disable_keep_power) >> + mmc->pm_caps |= MMC_PM_KEEP_POWER; >> >> /* >> * We can do SGIO >> diff --git a/drivers/mmc/host/mmci.h b/drivers/mmc/host/mmci.h >> index e1a9b96a3396..2cad1ef9766a 100644 >> --- a/drivers/mmc/host/mmci.h >> +++ b/drivers/mmc/host/mmci.h >> @@ -361,6 +361,7 @@ struct variant_data { >> u32 opendrain; >> u8 dma_lli:1; >> u32 stm32_idmabsize_mask; >> + u8 disable_keep_power:1; > > There are already four separate bitfields in struct variant_data, why > not move this up into one of them? > > regards > Philipp Hi Philipp, You're right, I'll put it with the previous bitfiled in v3. Best regards, Yann ^ permalink raw reply [flat|nested] 14+ messages in thread
end of thread, other threads:[~2022-03-15 12:31 UTC | newest] Thread overview: 14+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2022-03-14 9:52 [PATCH RESEND] mmc: mmci: manage MMC_PM_KEEP_POWER per variant config Yann Gautier 2022-03-14 11:17 ` Ulf Hansson 2022-03-14 12:20 ` Yann Gautier 2022-03-14 12:53 ` Ulf Hansson 2022-03-14 12:55 ` [PATCH v2] " Yann Gautier 2022-03-14 13:03 ` Ulf Hansson 2022-03-14 14:34 ` Yann Gautier 2022-03-14 16:18 ` Ulf Hansson 2022-03-14 16:51 ` Yann Gautier 2022-03-15 9:07 ` Ulf Hansson 2022-03-15 10:07 ` Yann Gautier 2022-03-15 12:30 ` Ulf Hansson 2022-03-14 13:05 ` Philipp Zabel 2022-03-14 13:45 ` Yann Gautier
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).