soc: qcom: llcc: Rename llcc-sdm845 to llcc-plat
Subsystems (2)
PATCH 1/2 soc: qcom: llcc: Rename llcc-sdm845 to llcc-plat
To avoid adding files for each future supported SoCs rename the file to a generic name - llcc-plat, so that llcc configuration tables for other SoCs can be added in the same driver.
---
drivers/soc/qcom/Kconfig | 10 +++++-----
drivers/soc/qcom/Makefile | 2 +-
drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} | 0
3 files changed, 6 insertions(+), 6 deletions(-)
rename drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} (100%)drivers/soc/qcom/Kconfig +5 -5
@@ -62,13 +62,13 @@ config QCOM_LLCC to clients that use the LLCC. Say yes here to enable LLCC slice driver. -config QCOM_SDM845_LLCC - tristate "Qualcomm Technologies, Inc. SDM845 LLCC driver" +config QCOM_PLAT_LLCC + tristate "Qualcomm Technologies, Inc. platform LLCC driver" depends on QCOM_LLCC help - Say yes here to enable the LLCC driver for SDM845. This provides - data required to configure LLCC so that clients can start using the - LLCC slices. + Say yes here to enable the LLCC driver for Qcom platforms, such as + SDM845. This provides data required to configure LLCC so that + clients can start using the LLCC slices. config QCOM_MDT_LOADER tristate
drivers/soc/qcom/Makefile +1 -1
@@ -21,6 +21,6 @@ obj-$(CONFIG_QCOM_SMSM) += smsm.o obj-$(CONFIG_QCOM_WCNSS_CTRL) += wcnss_ctrl.o obj-$(CONFIG_QCOM_APR) += apr.o obj-$(CONFIG_QCOM_LLCC) += llcc-slice.o -obj-$(CONFIG_QCOM_SDM845_LLCC) += llcc-sdm845.o +obj-$(CONFIG_QCOM_PLAT_LLCC) += llcc-plat.o obj-$(CONFIG_QCOM_RPMHPD) += rpmhpd.o obj-$(CONFIG_QCOM_RPMPD) += rpmpd.o
drivers/soc/qcom/llcc-plat.c renamed from drivers/soc/qcom/llcc-sdm845.c +0 -0
Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
PATCH 2/2 soc: qcom: llcc-plat: Make the driver more generic
- Remove 'sdm845' from names, and use 'plat' instead.
- Move SCT_ENTRY macro to header file.
- Create a new config structure to asssign to of-match-data.
--- drivers/soc/qcom/llcc-plat.c | 77 ++++++++++++-------------------------- include/linux/soc/qcom/llcc-qcom.h | 45 ++++++++++++++++++++++ 2 files changed, 68 insertions(+), 54 deletions(-)
drivers/soc/qcom/llcc-plat.c +23 -54
@@ -1,6 +1,6 @@// SPDX-License-Identifier: GPL-2.0 /* -* Copyright (c) 2017-2018, The Linux Foundation. All rights reserved. +* Copyright (c) 2017-2019, The Linux Foundation. All rights reserved. * */
@@ -10,47 +10,7 @@#include <linux/of_device.h> #include <linux/soc/qcom/llcc-qcom.h> -/* -* SCT(System Cache Table) entry contains of the following members: -* usecase_id: Unique id for the client's use case -* slice_id: llcc slice id for each client -* max_cap: The maximum capacity of the cache slice provided in KB -* priority: Priority of the client used to select victim line for replacement -* fixed_size: Boolean indicating if the slice has a fixed capacity -* bonus_ways: Bonus ways are additional ways to be used for any slice, -* if client ends up using more than reserved cache ways. Bonus -* ways are allocated only if they are not reserved for some -* other client. -* res_ways: Reserved ways for the cache slice, the reserved ways cannot -* be used by any other client than the one its assigned to. -* cache_mode: Each slice operates as a cache, this controls the mode of the -* slice: normal or TCM(Tightly Coupled Memory) -* probe_target_ways: Determines what ways to probe for access hit. When -* configured to 1 only bonus and reserved ways are probed. -* When configured to 0 all ways in llcc are probed. -* dis_cap_alloc: Disable capacity based allocation for a client -* retain_on_pc: If this bit is set and client has maintained active vote -* then the ways assigned to this client are not flushed on power -* collapse. -* activate_on_init: Activate the slice immediately after the SCT is programmed -*/ -#define SCT_ENTRY (uid, sid, mc, p, fs, bway, rway, cmod, ptw, dca, rp, a) \ - { \ - .usecase_id = uid, \ - .slice_id = sid, \ - .max_cap = mc, \ - .priority = p, \ - .fixed_size = fs, \ - .bonus_ways = bway, \ - .res_ways = rway, \ - .cache_mode = cmod, \ - .probe_target_ways = ptw, \ - .dis_cap_alloc = dca, \ - .retain_on_pc = rp, \ - .activate_on_init = a, \ - } - -static struct llcc_slice_config sdm845_data[] = { +static const struct llcc_slice_config sdm845_data[] = {SCT_ENTRY (LLCC_CPUSS ,1 ,2816 ,1 ,0 ,0xffc ,0x2 ,0 ,0 ,1 ,1 ,1 ),SCT_ENTRY (LLCC_VIDSC0 ,2 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),SCT_ENTRY (LLCC_VIDSC1 ,3 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),
@@ -71,30 +31,39 @@ static struct llcc_slice_config sdm845_data[] = { SCT_ENTRY(LLCC_AUDHW, 22, 1024, 1, 1, 0xffc, 0x2, 0, 0, 1, 1, 0), }; -static int sdm845_qcom_llcc_remove (struct platform_device *pdev) +static const struct qcom_llcc_config sdm845_cfg = { + .sct_data = sdm845_data, + .size =ARRAY_SIZE (sdm845_data), +}; + +static int qcom_plat_llcc_remove (struct platform_device *pdev) {return qcom_llcc_remove (pdev); } -static int sdm845_qcom_llcc_probe (struct platform_device *pdev) +static int qcom_plat_llcc_probe (struct platform_device *pdev) { -return qcom_llcc_probe (pdev, sdm845_data,ARRAY_SIZE (sdm845_data)); +const struct qcom_llcc_config *cfg; + + cfg =of_device_get_match_data (&pdev->dev); + +return qcom_llcc_probe (pdev, cfg->sct_data, cfg->size); } -static const struct of_device_id sdm845_qcom_llcc_of_match[] = { - { .compatible ="qcom,sdm845-llcc" , }, +static const struct of_device_id qcom_plat_llcc_of_match[] = { + { .compatible ="qcom,sdm845-llcc" , .data = &sdm845_cfg }, { } }; -static struct platform_driver sdm845_qcom_llcc_driver = { +static struct platform_driver qcom_plat_llcc_driver = { .driver = { - .name ="sdm845-llcc" , - .of_match_table = sdm845_qcom_llcc_of_match, + .name ="qcom-plat-llcc" , + .of_match_table = qcom_plat_llcc_of_match, }, - .probe = sdm845_qcom_llcc_probe, - .remove = sdm845_qcom_llcc_remove, + .probe = qcom_plat_llcc_probe, + .remove = qcom_plat_llcc_remove, }; -module_platform_driver (sdm845_qcom_llcc_driver); +module_platform_driver (qcom_plat_llcc_driver); -MODULE_DESCRIPTION ("QCOM sdm845 LLCC driver" ); +MODULE_DESCRIPTION ("QCOM platform LLCC driver" );MODULE_LICENSE ("GPL v2" );
include/linux/soc/qcom/llcc-qcom.h +45 -0
57 lines not on this page · raw message
Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
Re: [PATCH 2/2] soc: qcom: llcc-plat: Make the driver more generic
Quoted 4 lines · Vivek Gautam
- Remove 'sdm845' from names, and use 'plat' instead.
- MoveSCT_ENTRYmacro to header file.
- Create a new config structure to asssign to of-match-data.
I interpret the intention of these two patches as that you want to add some new platform without having to create one llcc-xyz.c per platform.
If that's the case then the only user of this macro would be in plat.c, so I don't see a reason for moving it to the header file.
Quoted 19 lines
Signed-off-by Vivek Gautam <[email protected]>--- drivers/soc/qcom/llcc-plat.c | 77 ++++++++++++-------------------------- include/linux/soc/qcom/llcc-qcom.h | 45 ++++++++++++++++++++++ 2 files changed, 68 insertions(+), 54 deletions(-)drivers/soc/qcom/llcc-plat.c +1
-1@@ -1,6 +1,6 @@// SPDX-License-Identifier: GPL-2.0 /* -* Copyright (c) 2017-2018, The Linux Foundation. All rights reserved. +* Copyright (c) 2017-2019, The Linux Foundation. All rights reserved. * */
drivers/soc/qcom/llcc-plat.c line 14
Quoted 5 lines
drivers/soc/qcom/llcc-plat.c +0
-2@@ -10,47 +10,7 @@#include <linux/of_device.h> #include <linux/soc/qcom/llcc-qcom.h> -/* -* SCT(System Cache Table) entry contains of the following members:
Should have caught this during previous review, but this comment simply duplicates the kerneldoc for struct llcc_slice_config.
Quoted 13 lines
drivers/soc/qcom/llcc-plat.c +0
-13-* usecase_id: Unique id for the client's use case -* slice_id: llcc slice id for each client -* max_cap: The maximum capacity of the cache slice provided in KB -* priority: Priority of the client used to select victim line for replacement -* fixed_size: Boolean indicating if the slice has a fixed capacity -* bonus_ways: Bonus ways are additional ways to be used for any slice, -* if client ends up using more than reserved cache ways. Bonus -* ways are allocated only if they are not reserved for some -* other client. -* res_ways: Reserved ways for the cache slice, the reserved ways cannot -* be used by any other client than the one its assigned to. -* cache_mode: Each slice operates as a cache, this controls the mode of the -* slice: normal or TCM(Tightly Coupled Memory)
drivers/soc/qcom/llcc-plat.c line 37
Quoted 10 lines
drivers/soc/qcom/llcc-plat.c +0
-10-* probe_target_ways: Determines what ways to probe for access hit. When -* configured to 1 only bonus and reserved ways are probed. -* When configured to 0 all ways in llcc are probed. -* dis_cap_alloc: Disable capacity based allocation for a client -* retain_on_pc: If this bit is set and client has maintained active vote -* then the ways assigned to this client are not flushed on power -* collapse. -* activate_on_init: Activate the slice immediately after the SCT is programmed -*/ -#define SCT_ENTRY (uid, sid, mc, p, fs, bway, rway, cmod, ptw, dca, rp, a) \
This simply maps macro arguments 1:1 to struct members, there's no need for a macro for this.
Quoted 49 lines
drivers/soc/qcom/llcc-plat.c +14
-21- { \ - .usecase_id = uid, \ - .slice_id = sid, \ - .max_cap = mc, \ - .priority = p, \ - .fixed_size = fs, \ - .bonus_ways = bway, \ - .res_ways = rway, \ - .cache_mode = cmod, \ - .probe_target_ways = ptw, \ - .dis_cap_alloc = dca, \ - .retain_on_pc = rp, \ - .activate_on_init = a, \ - } - -static struct llcc_slice_config sdm845_data[] = { +static const struct llcc_slice_config sdm845_data[] = {SCT_ENTRY (LLCC_CPUSS ,1 ,2816 ,1 ,0 ,0xffc ,0x2 ,0 ,0 ,1 ,1 ,1 ),SCT_ENTRY (LLCC_VIDSC0 ,2 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),SCT_ENTRY (LLCC_VIDSC1 ,3 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),@@ -71,30 +31,39 @@ static struct llcc_slice_config sdm845_data[] = { SCT_ENTRY(LLCC_AUDHW, 22, 1024, 1, 1, 0xffc, 0x2, 0, 0, 1, 1, 0), }; -static int sdm845_qcom_llcc_remove (struct platform_device *pdev) +static const struct qcom_llcc_config sdm845_cfg = { + .sct_data = sdm845_data, + .size =ARRAY_SIZE (sdm845_data), +}; + +static int qcom_plat_llcc_remove (struct platform_device *pdev) {return qcom_llcc_remove (pdev); } -static int sdm845_qcom_llcc_probe (struct platform_device *pdev) +static int qcom_plat_llcc_probe (struct platform_device *pdev) { -return qcom_llcc_probe (pdev, sdm845_data,ARRAY_SIZE (sdm845_data)); +const struct qcom_llcc_config *cfg; + + cfg =of_device_get_match_data (&pdev->dev); + +return qcom_llcc_probe (pdev, cfg->sct_data, cfg->size); } -static const struct of_device_id sdm845_qcom_llcc_of_match[] = { - { .compatible ="qcom,sdm845-llcc" , }, +static const struct of_device_id qcom_plat_llcc_of_match[] = {
drivers/soc/qcom/llcc-plat.c static struct llcc_slice_config sdm845_data[] = { · line 60
Quoted 10 lines
drivers/soc/qcom/llcc-plat.c +3
-3+ { .compatible = "qcom,sdm845-llcc", .data = &sdm845_cfg }, { } }; -static struct platform_driver sdm845_qcom_llcc_driver = { +static struct platform_driver qcom_plat_llcc_driver = { .driver = { - .name ="sdm845-llcc" , - .of_match_table = sdm845_qcom_llcc_of_match, + .name ="qcom-plat-llcc" ,
With this being the "one and only llcc driver", why not making it "qcom_llcc"?
include/linux/soc/qcom/llcc-qcom.h
Quoted 14 lines
drivers/soc/qcom/llcc-plat.c +5
-4+ .of_match_table = qcom_plat_llcc_of_match, }, - .probe = sdm845_qcom_llcc_probe, - .remove = sdm845_qcom_llcc_remove, + .probe = qcom_plat_llcc_probe, + .remove = qcom_plat_llcc_remove, }; -module_platform_driver (sdm845_qcom_llcc_driver); +module_platform_driver (qcom_plat_llcc_driver); -MODULE_DESCRIPTION ("QCOM sdm845 LLCC driver" ); +MODULE_DESCRIPTION ("QCOM platform LLCC driver" );MODULE_LICENSE ("GPL v2" );include/linux/soc/qcom/llcc-qcom.h +0
-0
This file should be describing the public interface to the llcc, the private pieces is better kept in drivers/soc/qcom/llcc.h
But this patch makes me wonder if there's a need to split llcc-slice and llcc-plat (and have a header file to describe API between them) instead of just having one file.
Regards,
Bjorn
Quoted 66 lines
index
eb71a50b8afc..8776bb5d3891100644include/linux/soc/qcom/llcc-qcom.h +45
-0@@ -27,6 +27,46 @@#define LLCC_MDMPNG 21#define LLCC_AUDHW 22 +/* +* SCT(System Cache Table) entry contains of the following members: +* usecase_id: Unique id for the client's use case +* slice_id: llcc slice id for each client +* max_cap: The maximum capacity of the cache slice provided in KB +* priority: Priority of the client used to select victim line for replacement +* fixed_size: Boolean indicating if the slice has a fixed capacity +* bonus_ways: Bonus ways are additional ways to be used for any slice, +* if client ends up using more than reserved cache ways. Bonus +* ways are allocated only if they are not reserved for some +* other client. +* res_ways: Reserved ways for the cache slice, the reserved ways cannot +* be used by any other client than the one its assigned to. +* cache_mode: Each slice operates as a cache, this controls the mode of the +* slice: normal or TCM(Tightly Coupled Memory) +* probe_target_ways: Determines what ways to probe for access hit. When +* configured to 1 only bonus and reserved ways are probed. +* When configured to 0 all ways in llcc are probed. +* dis_cap_alloc: Disable capacity based allocation for a client +* retain_on_pc: If this bit is set and client has maintained active vote +* then the ways assigned to this client are not flushed on power +* collapse. +* activate_on_init: Activate the slice immediately after the SCT is programmed +*/ +#define SCT_ENTRY (uid, sid, mc, p, fs, bway, rway, cmod, ptw, dca, rp, a) \ + { \ + .usecase_id = uid, \ + .slice_id = sid, \ + .max_cap = mc, \ + .priority = p, \ + .fixed_size = fs, \ + .bonus_ways = bway, \ + .res_ways = rway, \ + .cache_mode = cmod, \ + .probe_target_ways = ptw, \ + .dis_cap_alloc = dca, \ + .retain_on_pc = rp, \ + .activate_on_init = a, \ + } +/** * llcc_slice_desc - Cache slice descriptor * @slice_id: llcc slice id @@ -67,6 +107,11 @@ struct llcc_slice_config {bool activate_on_init; }; +struct qcom_llcc_config { +const struct llcc_slice_config *sct_data; +int size; +}; +/** * llcc_drv_data - Data associated with the llcc driver * @regmap: regmap associated with the llcc device Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
Re: [PATCH 2/2] soc: qcom: llcc-plat: Make the driver more generic
Hi Bjorn,
Thanks for the review.
Quoted 10 lines · Bjorn Andersson
Quoted 4 lines · Vivek Gautam
- Remove 'sdm845' from names, and use 'plat' instead.
- MoveSCT_ENTRYmacro to header file.
- Create a new config structure to asssign to of-match-data.I interpret the intention of these two patches as that you want to add some new platform without having to create one
llcc-xyz.cper platform.
That's right. The intention is to avoid creating a new platform specific file.
Quoted 3 lines
If that's the case then the only user of this macro would be in
plat.c, so I don't see a reason for moving it to the header file.
Alright. Better to keep it in the driver file itself.
Quoted 23 lines
Quoted 23 lines
Signed-off-by Vivek Gautam <[email protected]>--- drivers/soc/qcom/llcc-plat.c | 77 ++++++++++++-------------------------- include/linux/soc/qcom/llcc-qcom.h | 45 ++++++++++++++++++++++ 2 files changed, 68 insertions(+), 54 deletions(-)drivers/soc/qcom/llcc-plat.c +1
-1@@ -1,6 +1,6 @@// SPDX-License-Identifier: GPL-2.0 /* -* Copyright (c) 2017-2018, The Linux Foundation. All rights reserved. +* Copyright (c) 2017-2019, The Linux Foundation. All rights reserved. * */ @@ -10,47 +10,7 @@#include <linux/of_device.h> #include <linux/soc/qcom/llcc-qcom.h>
Quoted 5 lines
drivers/soc/qcom/llcc-plat.c line 14
Quoted 2 lines
drivers/soc/qcom/llcc-plat.c +0
-2-/* -* SCT(System Cache Table) entry contains of the following members: Should have caught this during previous review, but this comment simply duplicates the kerneldoc for struct
llcc_slice_config.
Ok, i noticed it now. Will clean it up. I can remove this comment, and update the one for struct llcc_slice_config.
Quoted 25 lines
Quoted 14 lines
drivers/soc/qcom/llcc-plat.c +0
-13-* usecase_id: Unique id for the client's use case -* slice_id: llcc slice id for each client -* max_cap: The maximum capacity of the cache slice provided in KB -* priority: Priority of the client used to select victim line for replacement -* fixed_size: Boolean indicating if the slice has a fixed capacity -* bonus_ways: Bonus ways are additional ways to be used for any slice, -* if client ends up using more than reserved cache ways. Bonus -* ways are allocated only if they are not reserved for some -* other client. -* res_ways: Reserved ways for the cache slice, the reserved ways cannot -* be used by any other client than the one its assigned to. -* cache_mode: Each slice operates as a cache, this controls the mode of the -* slice: normal or TCM(Tightly Coupled Memory) drivers/soc/qcom/llcc-plat.c line 37
Quoted 10 lines
drivers/soc/qcom/llcc-plat.c +0
-10-* probe_target_ways: Determines what ways to probe for access hit. When -* configured to 1 only bonus and reserved ways are probed. -* When configured to 0 all ways in llcc are probed. -* dis_cap_alloc: Disable capacity based allocation for a client -* retain_on_pc: If this bit is set and client has maintained active vote -* then the ways assigned to this client are not flushed on power -* collapse. -* activate_on_init: Activate the slice immediately after the SCT is programmed -*/ -#define SCT_ENTRY (uid, sid, mc, p, fs, bway, rway, cmod, ptw, dca, rp, a) \
Quoted 2 lines
This simply maps macro arguments 1:1 to struct members, there's no need for a macro for this.
Sure, will remove the macro.
Quoted 54 lines
Quoted 54 lines
drivers/soc/qcom/llcc-plat.c +15
-21- { \ - .usecase_id = uid, \ - .slice_id = sid, \ - .max_cap = mc, \ - .priority = p, \ - .fixed_size = fs, \ - .bonus_ways = bway, \ - .res_ways = rway, \ - .cache_mode = cmod, \ - .probe_target_ways = ptw, \ - .dis_cap_alloc = dca, \ - .retain_on_pc = rp, \ - .activate_on_init = a, \ - } - -static struct llcc_slice_config sdm845_data[] = { +static const struct llcc_slice_config sdm845_data[] = {SCT_ENTRY (LLCC_CPUSS ,1 ,2816 ,1 ,0 ,0xffc ,0x2 ,0 ,0 ,1 ,1 ,1 ),SCT_ENTRY (LLCC_VIDSC0 ,2 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),SCT_ENTRY (LLCC_VIDSC1 ,3 ,512 ,2 ,1 ,0x0 ,0x0f0 ,0 ,0 ,1 ,1 ,0 ),@@ -71,30 +31,39 @@ static struct llcc_slice_config sdm845_data[] = { SCT_ENTRY(LLCC_AUDHW, 22, 1024, 1, 1, 0xffc, 0x2, 0, 0, 1, 1, 0), }; -static int sdm845_qcom_llcc_remove (struct platform_device *pdev) +static const struct qcom_llcc_config sdm845_cfg = { + .sct_data = sdm845_data, + .size =ARRAY_SIZE (sdm845_data), +}; + +static int qcom_plat_llcc_remove (struct platform_device *pdev) {return qcom_llcc_remove (pdev); } -static int sdm845_qcom_llcc_probe (struct platform_device *pdev) +static int qcom_plat_llcc_probe (struct platform_device *pdev) { -return qcom_llcc_probe (pdev, sdm845_data,ARRAY_SIZE (sdm845_data)); +const struct qcom_llcc_config *cfg; + + cfg =of_device_get_match_data (&pdev->dev); + +return qcom_llcc_probe (pdev, cfg->sct_data, cfg->size); } -static const struct of_device_id sdm845_qcom_llcc_of_match[] = { - { .compatible ="qcom,sdm845-llcc" , }, +static const struct of_device_id qcom_plat_llcc_of_match[] = { + { .compatible ="qcom,sdm845-llcc" , .data = &sdm845_cfg }, { } };
Quoted 9 lines
drivers/soc/qcom/llcc-plat.c static struct llcc_slice_config sdm845_data[] = { · line 60
Quoted 6 lines
drivers/soc/qcom/llcc-plat.c +2
-3-static struct platform_driver sdm845_qcom_llcc_driver = { +static struct platform_driver qcom_plat_llcc_driver = { .driver = { - .name ="sdm845-llcc" , - .of_match_table = sdm845_qcom_llcc_of_match, + .name ="qcom-plat-llcc" ,With this being the "one and only llcc driver", why not making it "
qcom_llcc"?
Sure, will make it just "qcom_llcc" driver.
Quoted 11 lines
Quoted 11 lines
drivers/soc/qcom/llcc-plat.c +4
-3+ .of_match_table = qcom_plat_llcc_of_match, }, - .probe = sdm845_qcom_llcc_probe, - .remove = sdm845_qcom_llcc_remove, + .probe = qcom_plat_llcc_probe, + .remove = qcom_plat_llcc_remove, }; -module_platform_driver (sdm845_qcom_llcc_driver); +module_platform_driver (qcom_plat_llcc_driver);
Quoted 7 lines
include/linux/soc/qcom/llcc-qcom.h
Quoted 4 lines
drivers/soc/qcom/llcc-plat.c +1
-1-MODULE_DESCRIPTION ("QCOM sdm845 LLCC driver" ); +MODULE_DESCRIPTION ("QCOM platform LLCC driver" );MODULE_LICENSE ("GPL v2" );include/linux/soc/qcom/llcc-qcom.h +0
-0This file should be describing the public interface to the llcc, the private pieces is better kept in
drivers/soc/qcom/llcc.h
Yes. I will split the things in two separate files as suggested.
Quoted 4 lines
But this patch makes me wonder if there's a need to split llcc-slice and llcc-plat (and have a header file to describe API between them) instead of just having one file.
Nice opportunity to merge the two files? :)
Best regards
Vivek
Quoted 58 lines
Regards,
BjornQuoted 54 lines
index
eb71a50b8afc..8776bb5d3891100644include/linux/soc/qcom/llcc-qcom.h +40
-0@@ -27,6 +27,46 @@#define LLCC_MDMPNG 21#define LLCC_AUDHW 22 +/* +* SCT(System Cache Table) entry contains of the following members: +* usecase_id: Unique id for the client's use case +* slice_id: llcc slice id for each client +* max_cap: The maximum capacity of the cache slice provided in KB +* priority: Priority of the client used to select victim line for replacement +* fixed_size: Boolean indicating if the slice has a fixed capacity +* bonus_ways: Bonus ways are additional ways to be used for any slice, +* if client ends up using more than reserved cache ways. Bonus +* ways are allocated only if they are not reserved for some +* other client. +* res_ways: Reserved ways for the cache slice, the reserved ways cannot +* be used by any other client than the one its assigned to. +* cache_mode: Each slice operates as a cache, this controls the mode of the +* slice: normal or TCM(Tightly Coupled Memory) +* probe_target_ways: Determines what ways to probe for access hit. When +* configured to 1 only bonus and reserved ways are probed. +* When configured to 0 all ways in llcc are probed. +* dis_cap_alloc: Disable capacity based allocation for a client +* retain_on_pc: If this bit is set and client has maintained active vote +* then the ways assigned to this client are not flushed on power +* collapse. +* activate_on_init: Activate the slice immediately after the SCT is programmed +*/ +#define SCT_ENTRY (uid, sid, mc, p, fs, bway, rway, cmod, ptw, dca, rp, a) \ + { \ + .usecase_id = uid, \ + .slice_id = sid, \ + .max_cap = mc, \ + .priority = p, \ + .fixed_size = fs, \ + .bonus_ways = bway, \ + .res_ways = rway, \ + .cache_mode = cmod, \ + .probe_target_ways = ptw, \ + .dis_cap_alloc = dca, \ + .retain_on_pc = rp, \ + .activate_on_init = a, \ + } +/** * llcc_slice_desc - Cache slice descriptor * @slice_id: llcc slice id @@ -67,6 +107,11 @@ struct llcc_slice_config {bool activate_on_init; };
Quoted 12 lines
Quoted 9 lines
include/linux/soc/qcom/llcc-qcom.h +5
-1+struct qcom_llcc_config { +const struct llcc_slice_config *sct_data; +int size; +}; +/** * llcc_drv_data - Data associated with the llcc driver * @regmap: regmap associated with the llcc device -- Quoted 3 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
--
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
Quoted 4 lines · Vivek Gautam
To avoid adding files for each future supported SoCs rename the file to a generic name - llcc-plat, so that llcc configuration tables for other SoCs can be added in the same driver.
We've had a generic LLCC Kconfig option and then a specific SDM845 one, with this change we have two different generic options and both would either always be enabled or disabled.
So I think you should drop QCOM_SDM845_LLCC and build both llcc-slice and llcc-plat into the same qcom_llcc.ko instead.
Regards,
Bjorn
Quoted 51 lines
Signed-off-by Vivek Gautam <[email protected]>--- drivers/soc/qcom/Kconfig | 10 +++++----- drivers/soc/qcom/Makefile | 2 +- drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} | 0 3 files changed, 6 insertions(+), 6 deletions(-) rename drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} (100%)drivers/soc/qcom/Kconfig +5
-5@@ -62,13 +62,13 @@ config QCOM_LLCC to clients that use the LLCC. Say yes here to enable LLCC slice driver. -config QCOM_SDM845_LLCC - tristate "Qualcomm Technologies, Inc. SDM845 LLCC driver" +config QCOM_PLAT_LLCC + tristate "Qualcomm Technologies, Inc. platform LLCC driver" depends on QCOM_LLCC help - Say yes here to enable the LLCC driver for SDM845. This provides - data required to configure LLCC so that clients can start using the - LLCC slices. + Say yes here to enable the LLCC driver for Qcom platforms, such as + SDM845. This provides data required to configure LLCC so that + clients can start using the LLCC slices. config QCOM_MDT_LOADER tristatedrivers/soc/qcom/Makefile +1
-1@@ -21,6 +21,6 @@ obj-$(CONFIG_QCOM_SMSM) += smsm.o obj-$(CONFIG_QCOM_WCNSS_CTRL) += wcnss_ctrl.o obj-$(CONFIG_QCOM_APR) += apr.o obj-$(CONFIG_QCOM_LLCC) += llcc-slice.o -obj-$(CONFIG_QCOM_SDM845_LLCC) += llcc-sdm845.o +obj-$(CONFIG_QCOM_PLAT_LLCC) += llcc-plat.o obj-$(CONFIG_QCOM_RPMHPD) += rpmhpd.o obj-$(CONFIG_QCOM_RPMPD) += rpmpd.odrivers/soc/qcom/llcc-plat.c renamed from drivers/soc/qcom/llcc-sdm845.c +0
-0Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
Quoted 8 lines · Bjorn Andersson
Quoted 4 lines · Vivek Gautam
To avoid adding files for each future supported SoCs rename the file to a generic name - llcc-plat, so that llcc configuration tables for other SoCs can be added in the same driver.
Quoted 6 lines
We've had a generic LLCC
Kconfigoption and then a specific SDM845 one, with this change we have two different generic options and both would either always be enabled or disabled.So I think you should drop
QCOM_SDM845_LLCCand build both llcc-slice and llcc-plat into the same qcom_llcc.ko instead.
Yea. I can chuck off the llcc-slice module. But for readability would it still be
better to maintain separate files. I will drop the SDM845 config, and keep only QCOM_LLC.
Best regards
Vivek
Quoted 55 lines
Regards,
BjornQuoted 51 lines
Signed-off-by Vivek Gautam <[email protected]>--- drivers/soc/qcom/Kconfig | 10 +++++----- drivers/soc/qcom/Makefile | 2 +- drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} | 0 3 files changed, 6 insertions(+), 6 deletions(-) rename drivers/soc/qcom/{llcc-sdm845.c => llcc-plat.c} (100%)drivers/soc/qcom/Kconfig +5
-5@@ -62,13 +62,13 @@ config QCOM_LLCC to clients that use the LLCC. Say yes here to enable LLCC slice driver. -config QCOM_SDM845_LLCC - tristate "Qualcomm Technologies, Inc. SDM845 LLCC driver" +config QCOM_PLAT_LLCC + tristate "Qualcomm Technologies, Inc. platform LLCC driver" depends on QCOM_LLCC help - Say yes here to enable the LLCC driver for SDM845. This provides - data required to configure LLCC so that clients can start using the - LLCC slices. + Say yes here to enable the LLCC driver for Qcom platforms, such as + SDM845. This provides data required to configure LLCC so that + clients can start using the LLCC slices. config QCOM_MDT_LOADER tristatedrivers/soc/qcom/Makefile +1
-1@@ -21,6 +21,6 @@ obj-$(CONFIG_QCOM_SMSM) += smsm.o obj-$(CONFIG_QCOM_WCNSS_CTRL) += wcnss_ctrl.o obj-$(CONFIG_QCOM_APR) += apr.o obj-$(CONFIG_QCOM_LLCC) += llcc-slice.o -obj-$(CONFIG_QCOM_SDM845_LLCC) += llcc-sdm845.o +obj-$(CONFIG_QCOM_PLAT_LLCC) += llcc-plat.o obj-$(CONFIG_QCOM_RPMHPD) += rpmhpd.o obj-$(CONFIG_QCOM_RPMPD) += rpmpd.odrivers/soc/qcom/llcc-plat.c renamed from drivers/soc/qcom/llcc-sdm845.c +0
-1--QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation