Commit Graph

531 Commits

Author SHA1 Message Date
Ciro Santilli
5698a5ed2b build: move br2_cli.gitignore to out/ 2018-04-18 14:06:48 +01:00
Ciro Santilli
f4eed4163b scripts: factor out benchmarks and run further with ./eeval and common_bench_cmd 2018-04-18 12:47:43 +01:00
Ciro Santilli
96090a4c24 readme: aarch64 gem5 hpi slow: document gem5 update and cache size increase not working 2018-04-17 22:54:55 +01:00
Ciro Santilli
cc6099dd6b gem5: update to 2a9573f5942b5416fb0570cf5cb6cdecba733392 2018-04-17 22:54:21 +01:00
Ciro Santilli
ad6e311a8b readme: gem5 arm possible good panics... 2018-04-17 21:33:24 +01:00
Ciro Santilli
991dd49748 readme: arm gem5 slow fix good commit 2018-04-17 17:15:43 +01:00
Ciro Santilli
3f1979c1b2 readme stress: expected outcome of ps 2018-04-17 15:08:42 +01:00
Ciro Santilli
fad1eb0a99 readme: too may try outs 2018-04-17 15:07:27 +01:00
Ciro Santilli
50f2518656 kernel_module, sample_package: correctly forward Buildroot -O and -g flags
readme: improve optimization documentation
2018-04-17 15:05:27 +01:00
Ciro Santilli
ffa430eef9 readme: typo 2018-04-17 13:16:33 +01:00
Ciro Santilli
02b433e3ee readme: move compiler optimization section under Buildroot 2018-04-17 13:11:18 +01:00
Ciro Santilli
383e1c3120 readme: ./run -f does not need a - in the argument 2018-04-17 13:05:26 +01:00
Ciro Santilli
6e8cf30849 stress 2018-04-17 11:51:13 +01:00
Ciro Santilli
d5de5c6324 ltp 2018-04-17 11:15:22 +01:00
Ciro Santilli
de9dba276b readme: document failed attempt to get newer uclibc to fix aarch64 proc_events 2018-04-17 10:44:03 +01:00
Ciro Santilli
4e42e29f6d readme: recommend better br2 path data/br2 2018-04-17 10:28:18 +01:00
Ciro Santilli
72e9fcc4b0 proc_events.c: comment out aarch64 2018-04-17 09:22:20 +01:00
Ciro Santilli
0111ca406b CONFIG_PROC_EVENTS 2018-04-16 13:38:04 +01:00
Ciro Santilli
d58d4f0796 readme: more detail to GDB step debug userland non-init without -d 2018-04-16 11:06:14 +01:00
Ciro Santilli
bfba11afdd run: add qemu options directly into run
readme: add ascii tmux split screenshot to preamble
2018-04-16 01:09:11 +01:00
Ciro Santilli
24cdcdc756 readme: document rdtsc 2018-04-16 00:48:10 +01:00
Ciro Santilli
6eae8bb052 readme: document arm lx-symbols failure 2018-04-15 22:59:50 +01:00
Ciro Santilli
51e31cdc29 x86_64: rdstc
arm: attempt pmccntr, but didn't work, TODO why: no QEMU support?
2018-04-15 14:10:21 +01:00
Ciro Santilli
ae780f6750 Two working methods for module_init GDB step debugging!
Not perfect, but doable.
2018-04-15 11:05:03 +01:00
Ciro Santilli
3cc33a8fe8 readme: find module load address with pr_debug
Post mortem stack to line with GDB.
2018-04-14 23:36:51 +01:00
Ciro Santilli
f33dab4a1c linux: pr_debug with CONFIG_DYNAMIC_DEBUG=y 2018-04-14 21:27:47 +01:00
Ciro Santilli
a08a87dc0f readme: more info on kernel panics, oops an backtraces.
myinsmod: use either finit or init
2018-04-14 19:11:13 +01:00
Ciro Santilli
7b0bd10c0b ring0 priviledged actions 2018-04-13 14:27:03 +01:00
Ciro Santilli
898e945e6c Merge branch 'master' of github.com:cirosantilli/linux-kernel-module-cheat 2018-04-13 13:41:12 +01:00
Ciro Santilli
d992ba4e6e Generalize module_version.c into module_info.c 2018-04-13 09:13:22 +01:00
Ciro Santilli
2427a322fc kstrto 2018-04-13 08:37:00 +01:00
Ciro Santilli
42e634ffd1 debugfs.c: extend example with toplevel file, fop and kstrtoull_from_user 2018-04-13 07:58:58 +01:00
Ciro Santilli
94fdb4e003 kernel config: aarch64 CONFIG_8139CP for rtl8139 which is required for qemu record and replay 2018-04-12 13:41:19 +01:00
Ciro Santilli
c1c517a1c7 readme: link to gem5 spectre meltdown poc 2018-04-12 11:20:58 +01:00
Ciro Santilli
c382f96ea3 zip-img: remove images file before regenerating it 2018-04-11 13:18:34 +01:00
Ciro Santilli
08abfb1e85 run: Ctrl+C works and gets passed to guest... at last!
readme: document graphics failures on arm and aarch64 further.
2018-04-11 01:08:30 +01:00
Ciro Santilli
4601aff344 zip-img: create, zips all QEMU images
build-all: make gem5 rebuild optional

rootfs_post_image_script: prevent qemu-img from generating trace files
2018-04-10 23:24:42 +01:00
Ciro Santilli
ca1ce45811 buildroot baseline: use cpio to find minimal image sizes 2018-04-10 14:56:45 +01:00
Ciro Santilli
b6dc20cb74 trace-boot: automate instruction count on arm
Clean old trace boot results with new automated results.
2018-04-10 03:38:24 +01:00
Ciro Santilli
fb317f4778 readme: document failed attempts at QEMU ARM shutdown automation 2018-04-09 11:46:35 +01:00
Ciro Santilli
046a3a3643 build-all: also build gem5, document further 2018-04-09 10:47:06 +01:00
Ciro Santilli
8b77fe703b readme: kgdb only working in x86 graphic mode
Not sure if it ever worked outside, but I guess it didn't.

Fix should be simple, but not important enough.
2018-04-08 18:16:35 +01:00
Ciro Santilli
9e79b9b06f Save .log output files to out/arch/
This is cleaner as it removes yet another .gitignore rule and allows
for one command per arch.

This was prompted by the recent Docker patch creating files as root,
which then breaks future host runs.

Also separate all arguments by newlines. This:

- makes it much easier to minimize when reporting QEMU bugs
- shows the full command on the screen despite line wrapping being turned off
  by one of the annoying emulators, to be investigated.

The commands are not yet pristine, because:

- there are some leading spaces
- options not sorted

but it is already a huge improvement, and those operations are now be easy from Vim.

The general strategy is:

- end every variable containing a command with a newline
- when expanding such variables, don't add an extra newline
2018-04-08 17:49:14 +01:00
Ciro Santilli
21499559eb readme: extend actual commands ran to report upstream bugs 2018-04-08 16:09:42 +01:00
Ciro Santilli
ef421c6a43 qemu rr: update status 2018-04-08 15:30:47 +01:00
Ciro Santilli
105c506ac6 sample-package: create 2018-04-08 12:49:12 +01:00
Ciro Santilli
c5f4d2289e readme: -monitor is what is killing out ctrl-c 2018-04-08 11:41:34 +01:00
Ciro Santilli
294a8c2424 docker: crate docker host setup
readme: standardize keyboard input notation to the format: Ctrl-C
2018-04-08 10:27:26 +01:00
Ciro Santilli
ecf2f3b29e build: use nproc instead of nproc - 2
Someone was trying to build on a 1 core machine and it broke the build.

The rationale for the -2 was to leave 2 cores for your to use your computer still,
but let us trust the OS scheduler a bit more.
2018-04-07 19:57:41 +01:00
Ciro Santilli
d5950cb499 readme: benchmark initial build is on p51 2018-04-07 18:50:40 +01:00