Allen comments

Results 51 comments of


                                            Allen

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

Hi @yosefe I have replaced the code and upgraded to 1.18.0-rc2, but I find that the problem still exists and there is a difference. Now the callback ucp_am_send_nbx on the...

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

``` (gdb) p iface->read_index_elem[0] $23 = { flags = 1 '\001', am_id = 26 '\032', length = 4193, desc = { seg_id = 9732102, seg_size = 6291456, offset = 5265344...

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

Hi @yosefe, I've got some new clues. I printed the FIFO data of the server through gdb, and found that the function uct_mm_iface_fifo_has_new_data was judged to be false, which caused...

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

Hi @yosefe I may already know the root cause of the problem. https://github.com/openucx/ucx/blob/9ce35d0545b9298d579c50deb8763f2346d0f577/src/uct/sm/mm/base/mm_ep.c#L375-L378 When the client crashes before line 378, the statement `elem->flags = elem_flags;` is not executed, which causes...

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

Hi @yosefe Thank you for your reply. If that's the case, then it seems that shared memory transport cannot be used in the production environment. It appears there is no...

When using shared memory communication, ucp_am_send_nbx hangs and callback not invoked

> [@ivanallen](https://github.com/ivanallen) Hi, I’m also interesting in the RPC UCX. Actually I’m doing a similar project. It’s your project based on P2300 execution? Maybe we could have more communication. Hi...

Under UCS_THREAD_MODE_SINGLE, pthread_spin_lock experiences severe contention.

> Hi [@ivanallen](https://github.com/ivanallen), Have you been able to try using v1.20.x in case this issue was already addressed in newer versions? Thank you @roiedanino . I have upgraded ucx to...

Under UCS_THREAD_MODE_SINGLE, pthread_spin_lock experiences severe contention.

> [@ivanallen](https://github.com/ivanallen), can you please share the program or provide a minimal reproducer so it will be easier for us to reproduce? Hi @roiedanino I used ucx_perftest to reproduce the...

Is CRC validation for attachments under consideration?

> > 因为协议是用户不能感知的，baidu_std的协议自身有很多字段，所以协议部分，让框架做是合理的，保证协议部分所有内容都计算了。而attachment，完全是给用户设置的，这也完全可以由用户计算。 > > 我觉得，可以理解成，我把数据“交”给rpc传输，rpc可以有能力“保证”交给rpc和最终rpc收到的数据是一样的。参考tcp的checksum，也不是只算自己头的checksum，payload也会算。 The Checksum of the TCP is calculated by taking into account the TCP Header, TCP body, and Pseudo IP header. 有一些场景的 attachment 是从入口处开始就算好了，如果...

Question about the implementation of `photon::mutex`

Adopting this mechanism (the brpc solution, and seemingly the pthread_mutex solution as well) in photon might lead to coroutine starvation. If there is only one vcpu, after each unlock, the...