- User Since
- Mar 29 2016, 8:32 PM (320 w, 5 d)
Apr 14 2022
FYI, wanted to add some info. of how to repro. the issue. Doing a nullfs mount with nocache set and then doing an untar into that mount would fail on missing directory. With this change, that doesn't happen.
Jul 16 2021
It would be good to note, that to find out if a child is attached to vmd, that they will have to look for the vmd device and not the bus. I have some uncommitted code that had to skip a check in the PCI hot plug code when in VMD mode. So I added code to walk the chain to see how it was attached. It would be good to have a comment here why this was done.
Mar 16 2021
Has this been testing on systems without UEFI BIOS? I'm not sure what happens if efirt is loaded on a non-UEFI system? Please include testing on non UEFI system and UEFI system. QEMU can be used to test in legacy BIOS mode. Please include testing results from both environments. FYI, the loader should set hint.smbios.0.mem in UEFI mode so that can be used a quick check on the address. It is good to verify the address is not in the BIOS range when testing. Some UEFI systems have a combined legacy and UEFI env.
Feb 4 2021
Jan 28 2021
Dec 13 2020
Dec 8 2020
Looks good to me :-) We should never hit this code and I had never hit it before. Per the bug report this came from LSI/Avago/Broadcom via their Thunderbolt driver update before mrsas. I never understood what the issue was and never implemented the else part. This is to deal with physical disks attached to the RAID versus a volume. I put the DJA XXX so if we ever hit it we might be able to figure it out. Never hit it in real use but rarely used syspd mode.
Oct 10 2019
Sep 17 2019
Add new line after sentence.
Sep 7 2019
Accidentally removed the part that chnaged the resources for the
bus. Adding it back in the attachment.
Sep 6 2019
Need to add bus_generic_attach, to probe children when not
loaded by devd.
Add VMD support to FreeBSD - updated
Sep 3 2019
Thanks for spotting the dup. from old work.
Aug 23 2019
Thanks for the feed back. If's kind of a strange driver.
Aug 16 2018
Aug 3 2018
Remove NULL check since free does that.
Get rid of the malloc check and spurious ';'.
Jul 25 2018
Jul 23 2018
- Updating D15683: Change image activator to check length of interpreter is less then MAXPATHLEN. #
- Enter a brief description of the changes included in this update.
- The first line is used as subject, next lines as comment. #
- If you intended to create a new revision, use:
- $ arc diff --create
Jun 6 2018
Feb 22 2017
Update to address comments so that writing a new variable works.
Use the sane attribute default and if the default is being used,
check for the old attribute file and use that. This could cause
an issue if the user tried to over ride the attribute but it matched
the sane default. This is probably low risk.
Oct 18 2016
Oct 14 2016
Oct 13 2016
Getting scheduled after 100µs should not be a problem, It won't
make any I/O fail. If it was less then 100µs then it would be a
problem. Taking longer on a busy machine should be a good
thing since then it working better with with other things. The
watchdog resets could take longer but I doubt this will be a
problem with that.
That test is exactly what I wanted. So it looks good to me. I assume the CPU
load also goes way down since DELAY is a CPU hog!
Is the performance the same versus DELAY? Yielding is something that I wanted to do and locally we
use pause in the kcs_loop which makes IPMI take a long time. We also have code to dump kernel back
traces into the SEL. In that mode things have to work in polled mode. I haven't tried to integrate that with