@khng I think with this change even shadowing those global variables can be easily caught.

Looks like I put the wrong review link in rG5270cc9757fb4e0c1303ec44c2602b75acf3806d

update diff

I will try :)

In D45197#1031659, @gshapiro wrote:

Just a quick note that the PR listed on this is incorrect. This fix addresses PR 278394.

Just a quick note that the PR listed on this is incorrect. This fix addresses PR 278394.

@manu if you are (still) happy with the updated versions, the ports/packages are coming as well (see D44945).
I think we are getting more "stable" now in between supported versions and "noise".
Also while this adds forward matching at least for Intel (the firmware flavor for sc does not exist as the firmware is not public yet), shipping it like this means that after a release we can add the firmware package and the driver and fwget will already support it.

Re-genertae them based on updated scripts.

Update firmware to 20240513 (I have locally updated the commit message).
Also sort out some iwlwifi bits (we try to find the best supported
firmware version if the latest does not exist; may not have made a difference
to the lastet review version?)

In D45215#1031630, @imp wrote:

What does nvmecontrol identify ndaX say here? What's the optimal I/O boundary? And maybe we should just have a knob to disable trying to use it.

nvmecontrol identify nda4 | grep Opt
Optimal I/O Boundary:        256 blocks

• remove unused variable

FYI, I did tweak a few style things when pushing: 1) I moved the new extern variable declaration down next to the one other extern variable in the header (and in general type definitions are first before externs), and in kern_linker.c I rewrapped a few lines to fit in 80 cols.

Wow. I'm surprised this bug went unnoticed for so long. Can you get the fix into 14.1 ?

What does nvmecontrol identify ndaX say here? What's the optimal I/O boundary? And maybe we should just have a knob to disable trying to use it.

• rebase to main

This skip is not necessary any longer.

In mountd.8, the capitalized flags are sorted before the lowercase flags. With those two changes, manpages LGTM

Changed the structure to remove extra FAM attribute.

Key findings out of quick codewalk over g_new_bio() callers
sys/dev/nvme/nvme_ns.c, the nvme(4) namespace driver, calls g_new_bio() in the heaviest manner.

• A lot of child bios are allocated in a burst.
• The child bios are always allocated in the non-blocking way by g_new_bio(), regardless from the bio commands.
• The child bios are allocated and freed within the nvme(4) namespace driver.

@gallatin @markj Could you please test this patch? I've also tested this patch on EC2 VMs with both 4K and 16K base pages, but their device mappings don't trigger any L1 or L2C demotions in the direct map.

Diff updated for libsys and other recent changes.

Thanks for taking the time to review it. Based on your feedback, I updated the patch.
And Yes, please help commit them.