IRC channel logs

<damo22>azert, no i think there was a change to gnumach that unwires memory allocated with vm_allocate_contiguous

<damo22>vm_allocate_contiguous: Add missing page unwiring after making the area non-pageable

<damo22>if the memory is allocated to a physical contiguous block and used for DMA, it should not be unwired right?

<damo22>otherwise it could be swapped out

<damo22>for (i = 0; i < vm_page_atop(size); i++) <<--- shouldnt that be i = vm_page_atop(size); i < npages ?

<damo22> vm_page_unwire(&pages[i]);

<damo22>to unwire the pages above the power of two size?

<damo22>hmm after applying that patch i get the assertion triggered when i kill rumpnet and start netdde

<damo22>maybe there is no way to release the memory after killing a process that calls vm_allocate_contiguous?

<damo22>i removed a memcpy from the rumpnet code

<damo22>per packet

<damo22>it still is slow, maxes out at 1MB/s

<damo22>i never get a buffer with more than one packet in it

<damo22>so it copies packets out one at a time

<youpi>damo22: non-pageable means not possibly swapped out

<youpi>and they wouldn't be swapped out unless you are *really* running out of memory

<damo22>well wiring down the whole task fixes the zero paddr

<gnucode>sneek later tell p4r4D0xum that I went ahead and added "You can build this file with:\n gcc -pthread -rt -o context-switch context-switch.c" to https://darnassus.sceen.net/~hurd-web/faq/context_switch/?updated youpi normally merges changes (that are good) within a week or sooner.

<sneek>Welcome back gnucode, you have 1 message!

<sneek>gnucode, p4r4D0xum says: Yeah that's what I meant, you need to add -pthread -rt flags or it wont compile, thanks I will check it out later on

<gnucode>damo22: congrats on networking! That's awesome!

2025-06-15.log