User Details
- User Since
- May 30 2017, 11:42 AM (483 w, 5 d)
Fri, Sep 4
Thu, Sep 3
Amazing!
Briefly tested:
# ifconfig vtnet0.20 create vlan 20 vlandev vtnet0 # sh ./jib addm test vtnet0.20 vtnet0.20bridge e0a_test e0b_test
I was not able to test it with jng:
# kldload ng_ether # ifconfig vtnet0.20 create vlan 20 vlandev vtnet0 # jail -c name=test host.hostname=test.home.arpa vnet persist # sh ./jng bridge test vtnet0.20 ngctl: send msg: Invalid argument # ngctl list There are 3 total nodes: Name: vtnet0 Type: ether ID: 00000001 Num hooks: 0 Name: vtnet0_20 Type: ether ID: 00000002 Num hooks: 0 Name: ngctl3288 Type: socket ID: 00000005 Num hooks: 0 # sh ./jng bridge test vtnet0 # The parent works as expected ng0_test
Wed, Sep 2
Thank you!
This landed as 533918ced28ccd3b6ce22c93e0b64deac40377a1. The revision was not automatically closed because the commit message is slightly different (uppercase "R" in remove).
Let's examine release's functions one by one:
- arm_create_user is handled by another function in NanoBSD.
- arm_setup_usb_otg is somewhat handled by another function, but it is missing a few things.
- arm64_setup_multicons is handled by another function in NanoBSD.
- arm_setup_fdt_overlays is OK.
- arm_setup_minimal_loader I think we can ignore this customization?
- arm_do_quirk is achieved using nanobsd_loader_conf_extra (only used by PINEBOOK).
Tue, Sep 1
Mon, Aug 31
This change should ideally be done with D58964.
This is a followup of 62d028dcac1a ("Switch BEAGLEBONE, CUBIEBOARD, and PANDABOARD to use the GENERIC kernel configuration.")
If there are no objections, I will commit this revision over the weekend.
Sat, Aug 29
Fri, Aug 28
I believe we got the acronym wrong, shouldn't the directory name be sbc (Single-Board Computer)?
create_esp_partition() and create_cidata_partition() share a lot of common code. I wonder if they can be abstracted out?
Please squash with D58964.
This revision should be squashed with D59041.
Thu, Aug 27
Wed, Aug 26
Since this is a legacy-breaking change (perhaps the most significant one), we should provide an easy migration path. For example, in the commit message, we can explain that users who want to keep using sector sizes could add an s (for sector) suffix to its current values when migrating to the default plan, this is a deliberate change from strsuftoll, which uses b (for block) for that same purpose.
Please see: https://github.com/Coderangshu/freebsd-src/pull/60, which contains a WIP of NanoBSD using ZFS (the ZFS part is not really relevant at the moment, it was just an experiment). It illustrates that the file gpt.sh should actually be named ufs.sh and that common functions inherent to a GPT partitioning scheme belong in defaults.sh, leaving filesystem-specific functions inside their respective files (ufs.sh, zfs.sh, cd9660.sh, msdos.sh, i.e., anything that makefs(8) supports).
Maybe something closer to:
nanobsd: Disable dumpdev
I don't think we should touch anything legacy, or at least avoid it as much as possible. In this context embedded can be considered legacy.
I think _xxx_pkg_db_dump_or_vacuum should be added sooner. In fact, I think this function should not be a _xxx_ function, as we will always want to verify the integrity of the pkg database before shipping our NanoBSD image, so I suggest "promoting" it to a regular pkg function (pkg_db_vacuum since it no longer dumps.) and including it in D58951.
I think the commit message in this revision also needs updating.
While I do agree with the change, I do not fully agree with the commit message.
For me, the reason we want to remove this step is because we want to ship a NanoBSD image with the flags set just like a normal FreeBSD installation (/var/empty for instance).
I think this commit can be squashed with D58951, I see no reason to keep this in a separate commit.
Tue, Aug 25
This revision is just a placeholder for D54187?
Do you think I can commit D52203 after you commit this update, or maybe incorporate those changes here?
The changes from freebsd-configure.sh (config.h) and freebsd-namespace.sh (ssh_namespace.h) are in a separate commit as usual?
I have this change in libssh's Makefile:
--- a/secure/lib/libssh/Makefile +++ b/secure/lib/libssh/Makefile @@ -35,7 +35,7 @@ SRCS= ${LIBOPENSSH_SRCS} \ libcrux-mlkem-mldsa.c ssh-mldsa-eddsa.c \ hmac.c ed25519.c ed25519-openssl.c \ kex.c kex-names.c kexdh.c kexgex.c kexecdh.c kexc25519.c \ - kexgexc.c kexgexs.c \ + kexgexc.c kexgexs.c kexmlkem768ecdh.c \ kexsntrup761x25519.c kexmlkem768x25519.c sntrup761.c kexgen.c \ sftp-realpath.c platform-pledge.c platform-tracing.c platform-misc.c \ sshbuf-io.c misc-agent.c ssherr-libcrypto.c
Mon, Aug 24
Fix the output in the README file as well.
It seems that this should be squashed into D58964.
Is this commit fixing something already in base?
If the answer is yes, then the commit message should carry the trailer:
Fixes: <git log -n1 --format:'%h ("%s")' GIT-COMMIT-HASH>If the answer is no, the it should be squashed with the commit that introduced this change in you revision stack, glancing over it, it looks like it should be D58964.
Please provide the full URL: https://github.com/freebsd/pkg/issues/2672 in the commit log.
Sun, Aug 23
At this point, I think this can be squashed into D58951.
The comment is somewhat strange (I understand: Uplink sets learnMac=0, which prevents both learning new MAC addresses and moving existing hosts).
Squash into D58964.
Style only revision -> D58931.
Fri, Aug 21
I'm confused with the commit message: "since pkg generates them dynamically and does not emit them into package manifests."
- Dynamic generation is intrinsic to pkg.
- Perhaps most importantly is why we are doing this. It is because pkg does not generate a valid mtree (METALOG). All these metalog-related _xxx_ issues should point to https://github.com/freebsd/pkg/issues/2672.
I think this should only do a vacuum (integrity check).
So far we have not been able to create deterministic pkgbase images.
Remove git conflict markers.
This commit should be squashed with D58959.
Thu, Aug 20
This revision should be moved to very early in the stack, as this is an option we should already have for NanoBSD.
I fail to understand the "matching the existing entropy settings for the read only root" part in the commit message though.
This revision should be squashed with D58951.
In fact, those revisions that start the commit message with: Fix... should be squashed accordingly to its root commit, unless they fix something already in base.
As a side note, remember that the MAINTAINER of NanoBSD is @imp, so he must be tagged as a reviewer. He happens to be auto-subscribed to all revisions, but it is better to explicitly add him as a reviewer, at least for the crucial ones.
Thank you!
I prefer that when I issue a zpool status command, I see a GPT label. Please let me know if this is not the best practice anymore.
Wed, Aug 19
Hi @emaste!
Can I land this patch? I saw you vendor-imported 10.5p1?
I would recommend squashing the majority of this commit with D58931, as the changes are mostly NFC (the leftovers).
_xxx_-prefixed functions are supposed to be removed at some point:
As explained in the last call, I would prefer if each _xxx_-prefixed function is introduced in a separate commit, with its associated pull request/bug report. In that way, we can simply revert that commit once the fix is implemented.
For example, the function _xxx_libarchive_mtree_bug() is no longer necessary in -CURRENT, but we still want it to land so we can MFC it into stable/15, after which we can just revert the commit that introduced it.
I would even add some witty comment at the beginning of the file along the lines of: # "Nothing is more permanent than a temporary solution" as encouragement, in lieu of # Functions in this file should be removed before submitting our changes upstream, as that ship has already sailed.
This patch is a fix/cleanup, and should be squashed with the introduction of its original functions.