,
registration of 72 GB of GPU memory (on a Grace Hopper) is done in about
ten seconds, not hours.
rte_extmem_register(ext_mem.buf_ptr, ext_mem.buf_len, NULL, ext_mem.buf_iova,
GPU_PAGE_SIZE);
Thanks, John Romein
On 05-10-2024 00:16, Stephen Hemminger wrote:
On Wed, 1 Nov 2023 22:21:16
it seems the rte_mempool_mem_iter() functionality is
completely broken for the pools with external memory, and that's why
mlx5 implemented the dedicated branch to handle their registration.
With best regards,
Slava
-Original Message-
From: Aaron Conole
Sent: Tuesday, October 10, 2023 5
On 26-09-2023 17:04, Aaron Conole wrote:
Okay - we typically don't use pull requests. If you and others are are
okay, I can take the patches and repost them to the ML from the pull
requests with a note indicating such.
Yes please; that would be helpful.
Thanks, John
Dear Elena, Aaron,
I hope you had a nice time after the DPDK workshop.
I was unable to solve the issues with our mailserver to submit a patch
with git sendmail. So I created a pull request:
https://github.com/DPDK/dpdk/pull/69
@Elena, could you please handle the pull request? I will submit a
getenv needs stdlib.h to be included.
Bugzilla ID: 1133
Fixes: 24c77594e08f ("gpu/cuda: map GPU memory with GDRCopy")
Signed-off-by: John Romein
---
drivers/gpu/cuda/gdrcopy.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/cuda/gdrcopy.c b/drivers/gpu/cuda/gdrco
5 matches
Mail list logo
