Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | [rootfs-stage31] Workaround buggy e1000e versions | Simon Rettberg | 2022-07-11 | 1 | -0/+4 |
| | |||||
* | [init/kexec/] More advanced network setup via KCL, could skip DHCP | Simon Rettberg | 2020-07-08 | 1 | -22/+21 |
| | | | | | | | | | | | | | | | | If enough information if provided via KCL, we can skip the initial DHCP request in stage 3.1. Currently we require IP address, subnet mask and DNS server to skip DHCP. If we extend the boot server to supply a fallback/default DNS server, we could even do without DNS which would work in almost all cases. kexec-reboot has been extended to make use of this feature, but can be extended even more to provide up to date values from the current system configuration. Currently, some of the values it provides to the new kernel will be cached values from stage 3.1. kexec-reboot also honors the ipxe menu id from the KCL if given, to properly download the matching menu entry, which will take updates to the entry into account (e.g. changed URLs to kernel/init). | ||||
* | [rootfs-stage31] Actually request hostname via DHCP... | Simon Rettberg | 2019-11-26 | 1 | -1/+1 |
| | |||||
* | [rootfs-stage31] Clean up network config, increase delay on retry | Simon Rettberg | 2019-09-03 | 1 | -14/+13 |
| | |||||
* | [rfs-stage31] We don't have less, use more instead | Steffen Ritter | 2019-04-26 | 1 | -1/+1 |
| | |||||
* | [rootfs-stage31/dhcpc-busybox] Back to old DHCP behavior | Simon Rettberg | 2019-03-13 | 1 | -5/+7 |
| | | | | | | Since the PXE spec abolished th practive of having the PXE ROM set the uid field of the DHCP request to the system's UUID, we also shouldn't be doing it. Make it a KCL option called dhcpuuid. | ||||
* | [rfs-stage31] Double network setup timeout (operstate) | Simon Rettberg | 2019-01-17 | 1 | -1/+1 |
| | |||||
* | [rfs-*] Use SMBIOS UUID as UID for DHCP requests | Simon Rettberg | 2018-07-18 | 1 | -1/+8 |
| | | | | | | This is what the PXE rom does, so by using the same uid we make sure that we keep the same address in case the client is running in a dynamic address pool. | ||||
* | [rfs-stage31] Ordering | Simon Rettberg | 2018-05-04 | 1 | -1/+2 |
| | |||||
* | [rfs-stage31] Fix race when syncing time, more network setup tweaks | Simon Rettberg | 2018-05-04 | 1 | -2/+8 |
| | |||||
* | [rfs-stage31] Bring br0 up even if we don't have an IP yet | Simon Rettberg | 2018-05-03 | 1 | -1/+1 |
| | |||||
* | [vmware12/rfs-stage31] Fix jumbo frame support (get config first, then read ↵ | Simon Rettberg | 2017-12-21 | 1 | -1/+0 |
| | | | | value) | ||||
* | [rfs-stage31/vmware] Support jumbo frames via SLX_JUMBO_FRAMES | Simon Rettberg | 2017-12-20 | 1 | -0/+1 |
| | |||||
* | [rfs-stage31] Honor NTP servers returned by DHCP for timesync; fallback to ↵ | Simon Rettberg | 2017-11-27 | 1 | -1/+1 |
| | | | | $SLX_NOW | ||||
* | [rfs-stage31] Set up bridges for all NICs found | Simon Rettberg | 2017-11-24 | 1 | -0/+11 |
| | |||||
* | [rootfs-stage31] Don't open debug shell if we're booting with splash screen | Simon Rettberg | 2017-02-09 | 1 | -3/+9 |
| | | | | | This probably means the end user booted the system, so we don't want them to have a way to open a root shell... | ||||
* | merge with latest dev version (tm-scripts commit ↵ | Jonathan Bauer | 2016-12-23 | 1 | -2/+9 |
| | | | | f5a59daf8d70a9027118292cd40b18c221897408) | ||||
* | merge with latest dev version | Sebastian | 2016-04-25 | 1 | -0/+84 |