Hi Antonio, Ken, Miklos,
Currently, we use a preprocessor flag `TFM_MULTI_CORE_TOPOLOGY` to comment the veneer sections in the templates in multi-core topology. Each time before building, we have to run the Python script to generate new link script/scatter file with veneer disabled, to replace the existing ones. It becomes more inconvenient as the number of developers and users on feature-twincpu branch grows.
As Chris proposed on https://review.trustedfirmware.org/c/trusted-firmware-m/+/1527, does it also make sense to directly update the "generated" linker script/scatter file as well, on feature-twincpu branch? `TFM_MULTI_CORE_TOPOLOGY` will be a common flag used in multi-core topology and will help resolve our urgent problem. If the final improvement solution is completed on master branch, we will update the feature branch accordingly when merging it back to master branch.
Please let me know if there is a better option for feature-twincpu branch. Thank you.
Best regards, Hu Ziji
-----Original Message----- From: TF-M tf-m-bounces@lists.trustedfirmware.org On Behalf Of Antonio De Angelis via TF-M Sent: Thursday, July 11, 2019 3:53 AM To: TF-M@lists.trustedfirmware.org; nd nd@arm.com Subject: Re: [TF-M] Common scatter files and templates
Hi Chris,
you are right, that file is autogenerated from the template file but both are kept under source control. The reason for this is that the autogenerated file is not created at build time but by manually running the tfm_parse_manifest_list.py, which has to be run every time something in the manifest is changed, and the resulting autogenerated file is committed under source control as well.
On the other hand, the build system could be modified to run the parsers at build time so that the autogenerated files wouldn't have to be stored in source control, and we could keep only the template. These two alternatives are both equally valid in my view, but if there is strong consensus for the other we can discuss.
Thanks, Antonio
________________________________ From: TF-M tf-m-bounces@lists.trustedfirmware.org on behalf of Christopher Brand via TF-M tf-m@lists.trustedfirmware.org Sent: 10 July 2019 19:50 To: TF-M@lists.trustedfirmware.org; Miklos Balint Subject: [TF-M] Common scatter files and templates
Can somebody please help me understand this? $ ls platform/ext/common/armclang/ tfm_common_s.sct tfm_common_s.sct.template $ ls platform/ext/common/gcc tfm_common_s.ld tfm_common_s.ld.template In both directories, both files are under source control, but the non-template files say that they're auto-generated: /*********** WARNING: This is an auto-generated file. Do not edit! ***********/
It's unusual to see both the source file and the artifact under source control.
It seems that they're generated by tools/tfm_parse_manifest_list.py, but that doesn't seem to be run as part of the build, so when is it run?
Thanks,
Chris
This message and any attachments may contain confidential information from Cypress or its subsidiaries. If it has been received in error, please advise the sender and immediately delete this message. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m
Hi David,
We are planning to propose a change on this part to generate files during building instead of pre-generate manually, and it needs to be done step by step so the first is about the partition and services.
So please go ahead with your change on feature branch and sync it with master when this proposed feature is ready for usage.
The name of preprocessors may have a change so let's talk about what name (TFM_MULTI_CORE_TOPOLOGY?) looks more proper when you are trying to merge back.
BR
-Ken
-----Original Message----- From: David Hu (Arm Technology China) David.Hu@arm.com Sent: Thursday, July 11, 2019 10:07 AM To: Antonio De Angelis Antonio.DeAngelis@arm.com; tf- m@lists.trustedfirmware.org; Ken Liu (Arm Technology China) Ken.Liu@arm.com; Miklos Balint Miklos.Balint@arm.com Cc: nd nd@arm.com Subject: RE: Common scatter files and templates
Hi Antonio, Ken, Miklos,
Currently, we use a preprocessor flag `TFM_MULTI_CORE_TOPOLOGY` to comment the veneer sections in the templates in multi-core topology. Each time before building, we have to run the Python script to generate new link script/scatter file with veneer disabled, to replace the existing ones. It becomes more inconvenient as the number of developers and users on feature-twincpu branch grows.
As Chris proposed on https://review.trustedfirmware.org/c/trusted-firmware- m/+/1527, does it also make sense to directly update the "generated" linker script/scatter file as well, on feature-twincpu branch? `TFM_MULTI_CORE_TOPOLOGY` will be a common flag used in multi-core topology and will help resolve our urgent problem. If the final improvement solution is completed on master branch, we will update the feature branch accordingly when merging it back to master branch.
Please let me know if there is a better option for feature-twincpu branch. Thank you.
Best regards, Hu Ziji
-----Original Message----- From: TF-M tf-m-bounces@lists.trustedfirmware.org On Behalf Of Antonio De Angelis via TF-M Sent: Thursday, July 11, 2019 3:53 AM To: TF-M@lists.trustedfirmware.org; nd nd@arm.com Subject: Re: [TF-M] Common scatter files and templates
Hi Chris,
you are right, that file is autogenerated from the template file but both are kept under source control. The reason for this is that the autogenerated file is not created at build time but by manually running the tfm_parse_manifest_list.py, which has to be run every time something in the manifest is changed, and the resulting autogenerated file is committed under source control as well.
On the other hand, the build system could be modified to run the parsers at build time so that the autogenerated files wouldn't have to be stored in source control, and we could keep only the template. These two alternatives are both equally valid in my view, but if there is strong consensus for the other we can discuss.
Thanks, Antonio
From: TF-M tf-m-bounces@lists.trustedfirmware.org on behalf of Christopher Brand via TF-M tf-m@lists.trustedfirmware.org Sent: 10 July 2019 19:50 To: TF-M@lists.trustedfirmware.org; Miklos Balint Subject: [TF-M] Common scatter files and templates
Can somebody please help me understand this? $ ls platform/ext/common/armclang/ tfm_common_s.sct tfm_common_s.sct.template $ ls platform/ext/common/gcc tfm_common_s.ld tfm_common_s.ld.template In both directories, both files are under source control, but the non-template files say that they're auto-generated: /*********** WARNING: This is an auto-generated file. Do not edit! ***********/
It's unusual to see both the source file and the artifact under source control.
It seems that they're generated by tools/tfm_parse_manifest_list.py, but that doesn't seem to be run as part of the build, so when is it run?
Thanks,
Chris
This message and any attachments may contain confidential information from Cypress or its subsidiaries. If it has been received in error, please advise the sender and immediately delete this message. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m
Hi Ken,
Sure. Thx a lot.
Best regards, Hu Ziji
-----Original Message----- From: Ken Liu (Arm Technology China) Ken.Liu@arm.com Sent: Thursday, July 11, 2019 12:55 PM To: David Hu (Arm Technology China) David.Hu@arm.com; Antonio De Angelis Antonio.DeAngelis@arm.com; tf-m@lists.trustedfirmware.org; Miklos Balint Miklos.Balint@arm.com Cc: nd nd@arm.com Subject: RE: Common scatter files and templates
Hi David,
We are planning to propose a change on this part to generate files during building instead of pre-generate manually, and it needs to be done step by step so the first is about the partition and services.
So please go ahead with your change on feature branch and sync it with master when this proposed feature is ready for usage.
The name of preprocessors may have a change so let's talk about what name (TFM_MULTI_CORE_TOPOLOGY?) looks more proper when you are trying to merge back.
BR
-Ken
-----Original Message----- From: David Hu (Arm Technology China) David.Hu@arm.com Sent: Thursday, July 11, 2019 10:07 AM To: Antonio De Angelis Antonio.DeAngelis@arm.com; tf- m@lists.trustedfirmware.org; Ken Liu (Arm Technology China) Ken.Liu@arm.com; Miklos Balint Miklos.Balint@arm.com Cc: nd nd@arm.com Subject: RE: Common scatter files and templates
Hi Antonio, Ken, Miklos,
Currently, we use a preprocessor flag `TFM_MULTI_CORE_TOPOLOGY` to comment the veneer sections in the templates in multi-core topology. Each time before building, we have to run the Python script to generate new link script/scatter file with veneer disabled, to replace the existing ones. It becomes more inconvenient as the number of developers and users on feature-twincpu branch grows.
As Chris proposed on https://review.trustedfirmware.org/c/trusted-firmware- m/+/1527, does it also make sense to directly update the "generated" linker script/scatter file as well, on feature-twincpu branch? `TFM_MULTI_CORE_TOPOLOGY` will be a common flag used in multi-core topology and will help resolve our urgent problem. If the final improvement solution is completed on master branch, we will update the feature branch accordingly when merging it back to master branch.
Please let me know if there is a better option for feature-twincpu branch. Thank you.
Best regards, Hu Ziji
-----Original Message----- From: TF-M tf-m-bounces@lists.trustedfirmware.org On Behalf Of Antonio De Angelis via TF-M Sent: Thursday, July 11, 2019 3:53 AM To: TF-M@lists.trustedfirmware.org; nd nd@arm.com Subject: Re: [TF-M] Common scatter files and templates
Hi Chris,
you are right, that file is autogenerated from the template file but both are kept under source control. The reason for this is that the autogenerated file is not created at build time but by manually running the tfm_parse_manifest_list.py, which has to be run every time something in the manifest is changed, and the resulting autogenerated file is committed under source control as well.
On the other hand, the build system could be modified to run the parsers at build time so that the autogenerated files wouldn't have to be stored in source control, and we could keep only the template. These two alternatives are both equally valid in my view, but if there is strong consensus for the other we can discuss.
Thanks, Antonio
From: TF-M tf-m-bounces@lists.trustedfirmware.org on behalf of Christopher Brand via TF-M tf-m@lists.trustedfirmware.org Sent: 10 July 2019 19:50 To: TF-M@lists.trustedfirmware.org; Miklos Balint Subject: [TF-M] Common scatter files and templates
Can somebody please help me understand this? $ ls platform/ext/common/armclang/ tfm_common_s.sct tfm_common_s.sct.template $ ls platform/ext/common/gcc tfm_common_s.ld tfm_common_s.ld.template In both directories, both files are under source control, but the non-template files say that they're auto-generated: /*********** WARNING: This is an auto-generated file. Do not edit! ***********/
It's unusual to see both the source file and the artifact under source control.
It seems that they're generated by tools/tfm_parse_manifest_list.py, but that doesn't seem to be run as part of the build, so when is it run?
Thanks,
Chris
This message and any attachments may contain confidential information from Cypress or its subsidiaries. If it has been received in error, please advise the sender and immediately delete this message. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you. -- TF-M mailing list TF-M@lists.trustedfirmware.org https://lists.trustedfirmware.org/mailman/listinfo/tf-m
tf-m@lists.trustedfirmware.org
-
David Hu (Arm Technology China) -
Ken Liu (Arm Technology China)