| Message ID | 20230426091211.21557-1-tinghan.shen@mediatek.com |
|---|---|
| Headers | show |
| Series | Add support for MT8195 SCP 2nd core | expand |
Hi Tinghan, On Wed, Apr 26, 2023 at 05:12:05PM +0800, Tinghan Shen wrote: > This is the preparation for probing multi-core SCP. The remoteproc > initialization flow is similar on cores and is reused to avoid > redundant code. > > The registers of config and l1tcm are shared for multi-core > SCP. Reuse the mapped addresses for all cores. > > Signed-off-by: Tinghan Shen <tinghan.shen@mediatek.com> > Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> > --- > drivers/remoteproc/mtk_scp.c | 69 ++++++++++++++++++++++++++---------- > 1 file changed, 50 insertions(+), 19 deletions(-) > > diff --git a/drivers/remoteproc/mtk_scp.c b/drivers/remoteproc/mtk_scp.c > index 2bf66b1a8d80..5e4982f4d5dc 100644 > --- a/drivers/remoteproc/mtk_scp.c > +++ b/drivers/remoteproc/mtk_scp.c > @@ -23,6 +23,13 @@ > #define MAX_CODE_SIZE 0x500000 > #define SECTION_NAME_IPI_BUFFER ".ipi_buffer" > > +struct mtk_scp_of_cluster { > + void __iomem *reg_base; > + void __iomem *l1tcm_base; > + size_t l1tcm_size; > + phys_addr_t l1tcm_phys; > +}; This is a good start. > + > /** > * scp_get() - get a reference to SCP. > * > @@ -855,10 +862,11 @@ static void scp_remove_rpmsg_subdev(struct mtk_scp *scp) > } > } > > -static int scp_probe(struct platform_device *pdev) > +static int scp_rproc_init(struct platform_device *pdev) > { > struct device *dev = &pdev->dev; > struct device_node *np = dev->of_node; > + struct mtk_scp_of_cluster *of_cluster = platform_get_drvdata(pdev); 1) Because of the work done in the next patch, I think a "struct mtk_scp_of_cluster *" should be given as a parameter to scp_rproc_init(). 2) I would rename of_cluster to scp_cluster. > struct mtk_scp *scp; > struct rproc *rproc; > struct resource *res; > @@ -879,6 +887,11 @@ static int scp_probe(struct platform_device *pdev) > scp->data = of_device_get_match_data(dev); > platform_set_drvdata(pdev, scp); > > + scp->reg_base = of_cluster->reg_base; > + scp->l1tcm_base = of_cluster->l1tcm_base; > + scp->l1tcm_size = of_cluster->l1tcm_size; > + scp->l1tcm_phys = of_cluster->l1tcm_phys; > + > res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "sram"); > scp->sram_base = devm_ioremap_resource(dev, res); > if (IS_ERR(scp->sram_base)) > @@ -888,24 +901,6 @@ static int scp_probe(struct platform_device *pdev) > scp->sram_size = resource_size(res); > scp->sram_phys = res->start; > > - /* l1tcm is an optional memory region */ > - res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "l1tcm"); > - scp->l1tcm_base = devm_ioremap_resource(dev, res); > - if (IS_ERR(scp->l1tcm_base)) { > - ret = PTR_ERR(scp->l1tcm_base); > - if (ret != -EINVAL) { > - return dev_err_probe(dev, ret, "Failed to map l1tcm memory\n"); > - } > - } else { > - scp->l1tcm_size = resource_size(res); > - scp->l1tcm_phys = res->start; > - } > - > - scp->reg_base = devm_platform_ioremap_resource_byname(pdev, "cfg"); > - if (IS_ERR(scp->reg_base)) > - return dev_err_probe(dev, PTR_ERR(scp->reg_base), > - "Failed to parse and map cfg memory\n"); > - > ret = scp->data->scp_clk_get(scp); > if (ret) > return ret; > @@ -957,6 +952,42 @@ static int scp_probe(struct platform_device *pdev) > return ret; > } > > +static int scp_probe(struct platform_device *pdev) > +{ > + struct device *dev = &pdev->dev; > + struct mtk_scp_of_cluster *of_cluster; > + struct resource *res; > + int ret; > + > + of_cluster = devm_kzalloc(&pdev->dev, sizeof(*of_cluster), GFP_KERNEL); > + if (!of_cluster) > + return -ENOMEM; > + > + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "cfg"); > + of_cluster->reg_base = devm_ioremap_resource(dev, res); > + if (IS_ERR(of_cluster->reg_base)) > + return dev_err_probe(dev, PTR_ERR(of_cluster->reg_base), > + "Failed to parse and map cfg memory\n"); > + > + /* l1tcm is an optional memory region */ > + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "l1tcm"); > + of_cluster->l1tcm_base = devm_ioremap_resource(dev, res); > + if (IS_ERR(of_cluster->l1tcm_base)) { > + ret = PTR_ERR(of_cluster->l1tcm_base); > + if (ret != -EINVAL) > + return dev_err_probe(dev, ret, "Failed to map l1tcm memory\n"); > + > + of_cluster->l1tcm_base = NULL; > + } else { > + of_cluster->l1tcm_size = resource_size(res); > + of_cluster->l1tcm_phys = res->start; > + } > + > + platform_set_drvdata(pdev, of_cluster); > + In scp_rproc_init() pdev->dev->driver_data is set to a *scp, which defeats the purpose of setting here. Once the driver data for @pdev is set, it doesn't change after that. > + return scp_rproc_init(pdev); Function scp_rproc_init() needs to return an *scp and that scp needs to be added to the list of SCPs in scp_cluster. > +} > + > static int scp_remove(struct platform_device *pdev) > { > struct mtk_scp *scp = platform_get_drvdata(pdev); This should be: struct mtk_scp_of_cluster *scp_cluster = platform_get_drvdata(pdev); And then iterate over the list of SCPs to remove each scp. That way things work the same way regardless of the amount of scp in the system. > -- > 2.18.0 >
The mediatek remoteproc driver currently only allows bringing up a single core SCP, e.g. MT8183. It also only bringing up the 1st core in SoCs with a dual-core SCP, e.g. MT8195. This series support to bring-up the 2nd core of the dual-core SCP. v9 -> v10: 1. move the global mtk_scp list into the platform device driver data structure 2. remove an unnecessary if() condition v8 -> v9: 1. initialize l1tcm_size/l1tcm_phys at patchset 05/11 2. rewrite patchset 06/11 to unify the flow and remove hacks v7 -> v8: 1. update the node name of mt8192 asurada SCP rpmsg subnode 2. squash register definitions into driver patches 3. initialize local variables on the declaration at patch v8 06/11 v6 -> v7: 1. merge the mtk_scp_cluster struct into the mtk_scp structure at the "Probe multi-core SCP" patch v5 -> v6: 1. move the mtk_scp_of_regs structure from mtk_common.h to mtk_scp.c 2. rename the SCP core 0 label from 'scp' to 'scp_c0' v4 -> v5: 1. move resource release actions to the platform driver remove operation 2. fix dual-core watchdog handling v3 -> v4: 1. change the representation of dual-core SCP in dts file and update SCP yaml 2. rewrite SCP driver to reflect the change of dts node 3. drop 'remove redundant call of rproc_boot for SCP' in v3 for further investigation v2 -> v3: 1. change the representation of dual-core SCP in dts file and update SCP yaml 2. rewrite SCP driver to reflect the change of dts node 3. add SCP core 1 node to mt8195.dtsi 4. remove redundant call of rproc_boot for SCP 5. refine IPI error message v1 -> v2: 1. update dt-binding property description 2. remove kconfig for scp dual driver 3. merge mtk_scp_dual.c and mtk_scp_subdev.c to mtk_scp.c Tinghan Shen (11): dt-bindings: remoteproc: mediatek: Improve the rpmsg subnode definition arm64: dts: mediatek: Update the node name of SCP rpmsg subnode dt-bindings: remoteproc: mediatek: Support MT8195 dual-core SCP remoteproc: mediatek: Add MT8195 SCP core 1 operations remoteproc: mediatek: Extract remoteproc initialization flow remoteproc: mediatek: Probe multi-core SCP remoteproc: mediatek: Control SCP core 1 by rproc subdevice remoteproc: mediatek: Setup MT8195 SCP core 1 SRAM offset remoteproc: mediatek: Handle MT8195 SCP core 1 watchdog timeout remoteproc: mediatek: Refine ipi handler error message arm64: dts: mediatek: mt8195: Add SCP 2nd core .../bindings/remoteproc/mtk,scp.yaml | 176 +++++++- .../arm64/boot/dts/mediatek/mt8183-kukui.dtsi | 2 +- .../boot/dts/mediatek/mt8192-asurada.dtsi | 2 +- .../boot/dts/mediatek/mt8195-cherry.dtsi | 6 +- arch/arm64/boot/dts/mediatek/mt8195.dtsi | 32 +- drivers/remoteproc/mtk_common.h | 33 ++ drivers/remoteproc/mtk_scp.c | 415 ++++++++++++++++-- 7 files changed, 604 insertions(+), 62 deletions(-)