Kazuhito Hagio

@lian-bo, a __percpu pointer allocated by `alloc_percpu()` is not within the range. ``` crash> mount.mnt_pcp ffff916f5500b480 mnt_pcp = 0x340b01007114, crash> sym __per_cpu_start __per_cpu_end 0 (D) __per_cpu_start 2d000 (D) __per_cpu_end ```...

> Is it ok to use the debuginfo from other builds? Unfortunately no. crash uses some gdb functions and commands internally to achieve "dis -s" command, it will cause reference...

The `e4e00000` is calculated using `phys_base`, which varies with machine, KASLR, etc. It looks like the page is not mapped on your machine.

No, the current crash does not support riscv64. Probably the gdb-7.6 that crash contains also does not support it in the first place. We have been working on rebasing gdb...

The crash-7.2.9 doesn't support the 5.10 kernel. It needs these two patches a5531b2 and 71e159c, so please use the latest master if possible.

``` read_diskdump: SEEK_ERROR: paddr/pfn: 203ff7bc0090/203ff7bc max_mapnr: 203fc000 ``` Hmm, it's strange that pfn is higher than max_mapnr.. * What is the memory size of the machine? (32TB?) * Is it...

crash's `struct` command is based on gdb's `ptype` command, what happens if you open the module with gdb? Does the module have its debuginfo certainly? ``` $ gdb netmap.ko ......

Not reproducible at my end. What is the crash version, kernel version, architecture and distribution? ``` crash> list packet_offload.list -H offload_base -s packet_offload ffffffffae02a8e0 struct packet_offload { type = 8,...

The gdb-7.6 that crash contains looks to have some code to handle `gnu` type compression, but as far as I've tested with `eu-elfcompress -t gnu` (not `gcc -gz=zlib`) it didn't...

Is the kernel self-built? The debuginfo (vmlinux) should have the source file paths, so I think you need to install the source files there, if it's a different machine from...