On Oct 27, 2021, at 3:55 AM, Richard Miller <9f...@hamnavoe.com> wrote: >> Skip, did you specify -o cache=mmap when mounting diod service >> for the go build experiment? > > I tried it myself using local diod and cache=mmap. I get a similar > SIGBUS on instruction fetch again. Conclusion: as Bakul says, now > I'm debugging linux. Not going there, thanks.
Looks like you are going there? :-) > Going further off-topic for 9fans, sorry: > > I thought it would be clever to update the linux client to a newer > kernel (4.19 was the latest I could find for debian 9). That > didn't go well: booting the new kernel fails with the message > Failed to find cpu0 device node > > Does anyone know if it's feasible to do an out-of-tree build > of v9fs kernel modules (9p, 9pnet?) from current source [where > is it?] and use them with my old 4.9 kernel? May be this will help? https://itnext.io/a-standalone-linux-kernel-module-df54283d4803 Though may you first want to start looking at the output of git log github.com/torvalds/linux/fs/9p/vfs_file.c grep for mmap related changes and check the code around changes to see if you can spot the bug. The last mmap mention in the log makes me wonder whether you see a difference in a 32 bit wordsize machine vs 64 bit. May also be worth just talking to the folks who worked on this feature. A long time ago I wrote an mmap example C program that works on anonymous maps (just steps through memory given some stride and allocates memory in the signal handler on page faults). Something similar can be written to access an mmapped file and tweaked until your bug appears. Currently my program compiles on freebsd but not linux (ubuntu). I will see if I can make it work. ------------------------------------------ 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tb065f4df67a8bab9-M30a0d7f933429868ec22fdd7 Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
