Skip to content

memory: removed hp buffer heap - #1567

Merged
tlauda merged 2 commits into
thesofproject:masterfrom
dabekjakub:heap
Oct 21, 2019
Merged

tlauda merged 2 commits into
thesofproject:masterfrom
dabekjakub:heap

Conversation

@dabekjakub

Copy link
Copy Markdown
Member

Merged hp buffer heap with normal buffer heap.
Since current buffer heap was actualy using hp memory and it served
no purpose to have two heaps.
Removed extern capabilities and merged other capabiliteis to
buffer heap. Currently none of those buffer heaps supports extern.

Signed-off-by: Jakub Dabek jakub.dabek@intel.com

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why such value?

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It is full bank size - aligment is easier.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dabekjakub Thanks for info

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why did it change from 0 to 0xBExxxxxx ?

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Because 0 is not realy a valid adress for a bootloader base.
Since HP heap is deleted and its base was used by bootloader as a stack we had to come up with a different place to store it. Hence the calculation.

Comment thread test/cmocka/src/lib/alloc/alloc.c Resolved

@abonislawski abonislawski left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

HEAP_BUFFER_SIZE not increased in APL?

@dabekjakub dabekjakub changed the title memory: removed hp buffer heap [DO NOT MERGE] memory: removed hp buffer heap Jun 14, 2019
@dabekjakub
dabekjakub force-pushed the heap branch 4 times, most recently from 720e196 to 67c7481 Compare June 22, 2019 05:41
@dabekjakub

Copy link
Copy Markdown
Member Author

@lgirdwood @xiulipan could you help me pinpoint the issue with quemu here ?i tryed booting up this fw on my boba platform and it works.

@lgirdwood

Copy link
Copy Markdown
Member

@xiulipan the qemu APL log is sending the FW ready message so not sure why its failing ?

@xiulipan

Copy link
Copy Markdown
Contributor

@dabekjakub
Is this PR change the MAILBOX offset? If so that maybe my scripts issue. Please also change the
OUTBOX_OFFSET="f000"
in scripts/qemu-check.sh with in the PR to make sure all CI follow your change.

@dabekjakub
dabekjakub force-pushed the heap branch 3 times, most recently from 92bc16a to bd21156 Compare June 25, 2019 12:52
@tlauda

tlauda commented Jun 26, 2019

Copy link
Copy Markdown
Contributor

@dabekjakub UT are failing.

@dabekjakub dabekjakub changed the title [DO NOT MERGE] memory: removed hp buffer heap memory: removed hp buffer heap Jun 26, 2019
@RanderWang

Copy link
Copy Markdown
Collaborator

@dabekjakub what is the status now ? I have a PR to unify the memory allocation and the PR is linked to your PR.

@lgirdwood

Copy link
Copy Markdown
Member

@dabekjakub ping, can you give us an ETA ?

@dabekjakub

dabekjakub commented Jul 31, 2019 •

Copy link
Copy Markdown
Member Author

@RanderWang @lgirdwood Actualy before this removal can be implemented memory alignment feature (that i just finished working on ) must be made.
Otherwise unaligned DMA buffers result in Glitches on playback.
I will link PR that needs to be merged later today.

@dabekjakub

Copy link
Copy Markdown
Member Author

#1694 Is the PR that has to come first so the DMA buffer will align correctly

@lgirdwood

lgirdwood commented Aug 2, 2019 •

Copy link
Copy Markdown
Member

@dabekjakub can you fix conflicts and then @mengdonglin can schedule someone to test.

@dabekjakub

Copy link
Copy Markdown
Member Author

@lgirdwood @mengdonglin It could use a round of back to back testing , since the buffer pointers were moved (the alignment commit should have fixed that, however, it is good to be sure).
@zrombel Could we get full regression on this before the merge?

@wenqingfu

Copy link
Copy Markdown

SOFCI TEST

@dabekjakub

Copy link
Copy Markdown
Member Author

@jajanusz do You think that the

#define DMA_BUFFER_ALIGNMENT 128

should stay in buffer.h or should it be platform-specific in memory.h?

Comment thread src/include/sof/audio/buffer.h Resolved
@dabekjakub
dabekjakub force-pushed the heap branch 5 times, most recently from cdb3374 to b007236 Compare October 18, 2019 11:47
Comment thread src/audio/host.c Outdated

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fix checkpatch.

Comment thread src/include/sof/audio/buffer.h Outdated

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please remove.

Merged hp buffer heap with normal buffer heap.
Since current buffer heap was actually using hp memory and it served
no purpose to have two heaps.
Removed extern capabilities and merged other capabiliteis to
buffer heap. Currently none of those buffer heaps supports extern.

Signed-off-by: Jakub Dabek <jakub.dabek@intel.com>
Add seeting of alignment for DMA buffer that is required by
Hardware configuration.

Signed-off-by: Jakub Dabek <jakub.dabek@intel.com>
@tlauda
tlauda merged commit 0234df5 into thesofproject:master Oct 21, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

8 participants