nvnetdrv v2 part 2 #689
Merged
nvnetdrv v2 part 2 #689
+71
−62
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
thrimbor
reviewed
Feb 21, 2025
disean
reviewed
Feb 22, 2025
| @@ -441,17 +447,25 @@ int nvnetdrv_init (size_t rx_buffer_count, nvnetdrv_rx_callback_t rx_callback) | |||
| } | |||
|
|
|||
| // Allocate memory for TX and RX ring descriptors. | |||
| void *descriptors = MmAllocateContiguousMemoryEx((RX_RING_SIZE + TX_RING_SIZE) * sizeof(struct descriptor_t), 0, | |||
| void *descriptors = MmAllocateContiguousMemoryEx((g_rxRingSize + g_txRingSize) * sizeof(struct descriptor_t), 0, | |||
There was a problem hiding this comment.
For performance reasons, you may want to align this pointer on a 64-byte boundary.
Suggested change
| void *descriptors = MmAllocateContiguousMemoryEx((g_rxRingSize + g_txRingSize) * sizeof(struct descriptor_t), 0, | |
| // Allocate memory for TX and RX ring descriptors + alignment. | |
| void *descriptors = MmAllocateContiguousMemoryEx((g_rxRingSize + g_txRingSize) * sizeof(struct descriptor_t) + (64 - 1), 0, | |
| // Round up to the closest multiple of the cache line boundary. | |
| descriptors = (void *)((size_t)descriptors + (64 - (size_t)descriptors) % 64); |
There was a problem hiding this comment.
MmAllocateContiguousMemoryEx should allocate physical memory pages, so I'm not sure if this could ever not be aligned to 0x1000 (which is a multiple of 0x40)?
(There's some other mallocs which could be aligned_alloc but I'm not sure if there's actual benefit to be expected here)
There was a problem hiding this comment.
Yeah it's always page-aligned. MmAllocateContiguousMemoryEx has an alignment parameter, but that only does something if you want a POT alignment bigger than 4KiB.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Continuation from #686
Will likely end up in 3 parts.
Primary change here are to pass ring sizes to the init function instead of #defines. Previously we could have different hardware ring size and queue size. Its actually much simpler to just have the hardware ring size exactly the same size as the queue. It also works better with LWIP as we can create pbuf pools to exactly match our requirement.
Most other commits are just some minor bug fixes and house keeping.
This fixes a bug that on soft reset or coming from XDK app where the network would not retrieve an IP address in some cases.