Page MenuHomeFreeBSD

scottph (D Scott Phillips)
User

Projects

User does not belong to any projects.

User Details

User Since
May 22 2018, 5:54 PM (122 w, 6 d)

Recent Activity

Mon, Sep 21

scottph committed rS365982: stand/reloc_elf: Handle relative relocations for arm{,64} and riscv.
stand/reloc_elf: Handle relative relocations for arm{,64} and riscv
Mon, Sep 21, 10:25 PM
scottph closed D26365: stand/reloc_elf: Handle relative relocations for arm{,64} and riscv.
Mon, Sep 21, 10:25 PM
scottph committed rS365981: arm64/pmap: Sparsify pv_table.
arm64/pmap: Sparsify pv_table
Mon, Sep 21, 10:24 PM
scottph closed D26132: arm64/pmap: Sparsify pv_table.
Mon, Sep 21, 10:24 PM
scottph committed rS365980: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE
Mon, Sep 21, 10:23 PM
scottph closed D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Mon, Sep 21, 10:23 PM
scottph committed rS365978: Sparsify the vm_page_dump bitmap.
Sparsify the vm_page_dump bitmap
Mon, Sep 21, 10:22 PM
scottph closed D26131: Sparsify the vm_page_dump bitmap.
Mon, Sep 21, 10:22 PM
scottph committed rS365977: Move vm_page_dump bitset array definition to MI code.
Move vm_page_dump bitset array definition to MI code
Mon, Sep 21, 10:21 PM
scottph closed D26129: Move vm_page_dump bitset array definition to MI code.
Mon, Sep 21, 10:21 PM
scottph closed D26190: bitset: expand bit index type to `long`.
Mon, Sep 21, 10:19 PM
scottph committed rS365975: bitset: expand bit index type to `long`.
bitset: expand bit index type to `long`
Mon, Sep 21, 10:19 PM
scottph committed rS365974: vchi: rename bitset macros to avoid collision with bitset(9).
vchi: rename bitset macros to avoid collision with bitset(9)
Mon, Sep 21, 10:18 PM
scottph closed D26177: vchi: rename bitset macros to avoid collision with bitset(9).
Mon, Sep 21, 10:18 PM
scottph abandoned D26519: arm64/pmap: Remove valid bit check from r365879 assertions.
Mon, Sep 21, 9:31 PM
scottph added a comment to D26519: arm64/pmap: Remove valid bit check from r365879 assertions.

See D26510

Mon, Sep 21, 9:31 PM
scottph requested review of D26519: arm64/pmap: Remove valid bit check from r365879 assertions.
Mon, Sep 21, 9:07 PM

Thu, Sep 10

scottph added a comment to D26177: vchi: rename bitset macros to avoid collision with bitset(9).

I wonder if we can upstream this too

Thu, Sep 10, 1:36 AM

Wed, Sep 9

scottph added a comment to D26365: stand/reloc_elf: Handle relative relocations for arm{,64} and riscv.
In D26365#586260, @manu wrote:

Thanks !
My investigation of this problem 1 year ago lead me to think it was relocation related but I'm clueless on how that work.
I'll try to test that asap (especially arm where I don't think we had a problem).

Wed, Sep 9, 6:05 PM
scottph committed rS365505: stand/efihttp: Work around a bug in edk2 http instance reconfiguration.
stand/efihttp: Work around a bug in edk2 http instance reconfiguration
Wed, Sep 9, 4:36 PM
scottph closed D21281: Work around a bug in the EFI HTTP driver.
Wed, Sep 9, 4:36 PM
scottph requested review of D26365: stand/reloc_elf: Handle relative relocations for arm{,64} and riscv.
Wed, Sep 9, 2:26 AM

Tue, Sep 8

scottph added a comment to D21281: Work around a bug in the EFI HTTP driver.

I've tracked this down to a particular use-after-free type bug in edk2: https://www.mail-archive.com/devel@edk2.groups.io/msg25730.html

Tue, Sep 8, 4:57 AM

Mon, Aug 31

scottph added a comment to D26131: Sparsify the vm_page_dump bitmap.
In D26131#582102, @jhb wrote:

In general it seems like there's an implicit assumption that nothing will ask
for a PA that isn't in dump_avail[], and if it does we just return the wrong
page since we just return some offset relative to the previous dump_avail
entry that overflows into the next entry's bitmap? It would be annoying to
fix as it means all the places that call _kvm_pa_to_bit(), and or adding some
kind of _kvm_has_pa().

Mon, Aug 31, 11:18 PM
scottph updated the diff for D26131: Sparsify the vm_page_dump bitmap.
Mon, Aug 31, 11:12 PM
scottph added inline comments to D26132: arm64/pmap: Sparsify pv_table.
Mon, Aug 31, 10:47 PM
scottph updated the diff for D26132: arm64/pmap: Sparsify pv_table.
Mon, Aug 31, 10:28 PM

Aug 27 2020

scottph added a comment to D26190: bitset: expand bit index type to `long`.

Can we reuse the algorithm from is_dumpable()? That is, if the PA corresponds
to a vm_page, use a flag to decide whether the page is to be included in the
minidump, otherwise include it so long as the PA is in dump_avail[].

Aug 27 2020, 5:50 PM
scottph added a comment to D26190: bitset: expand bit index type to `long`.
In D26190#582299, @kib wrote:

If dump bitmap becomes so large, may be it is better to get rid of it at all ?
You can use vm_page aflags to track dumpable state.

Aug 27 2020, 4:48 PM

Aug 26 2020

scottph added inline comments to D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 26 2020, 9:01 PM
scottph updated the diff for D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 26 2020, 8:52 PM
scottph requested review of D26190: bitset: expand bit index type to `long`.
Aug 26 2020, 6:42 AM
scottph added inline comments to D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 26 2020, 6:40 AM
scottph updated the diff for D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 26 2020, 6:33 AM
scottph added a comment to D25839: efibootmgr: Add option to request booting to the firmware user interface.

So this is the moral equivalent of ipmitool chassis bootdev bios?

Aug 26 2020, 2:17 AM
scottph committed rS364798: arm64: Increase dmap size to 95 TiB.
arm64: Increase dmap size to 95 TiB
Aug 26 2020, 2:13 AM
scottph closed D26134: arm64: Increase dmap size to 95 TiB.
Aug 26 2020, 2:13 AM
scottph committed rS364797: arm64/acpi: Give the real PA limit to ACPI.
arm64/acpi: Give the real PA limit to ACPI
Aug 26 2020, 2:12 AM
scottph closed D26133: arm64/acpi: Give the real PA limit to ACPI.
Aug 26 2020, 2:12 AM
scottph committed rS364796: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit
Aug 26 2020, 2:08 AM
scottph closed D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 26 2020, 2:08 AM
scottph closed D25839: efibootmgr: Add option to request booting to the firmware user interface.
Aug 26 2020, 2:06 AM
scottph committed rS364795: efibootmgr: Add option to request booting to the firmware user interface.
efibootmgr: Add option to request booting to the firmware user interface
Aug 26 2020, 2:06 AM
scottph committed rS364794: arm64: Make local stores observable before sending IPIs.
arm64: Make local stores observable before sending IPIs
Aug 26 2020, 2:04 AM
scottph closed D25798: arm64: Make local stores observable before sending IPIs.
Aug 26 2020, 2:04 AM

Aug 25 2020

scottph added a comment to D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
In D26130#581703, @alc wrote:

I'm curious to see the output from "sysctl vm.phys_segs" on this machine.

Aug 25 2020, 5:08 PM

Aug 24 2020

scottph added a comment to D26129: Move vm_page_dump bitset array definition to MI code.

Adding bitset.h to vm_page.h breaks the vhci driver. D26177 fixes that.

Aug 24 2020, 11:52 PM
scottph requested review of D26177: vchi: rename bitset macros to avoid collision with bitset(9).
Aug 24 2020, 11:50 PM
scottph added a comment to D26132: arm64/pmap: Sparsify pv_table.

I augmented vm_phys_seg with a void *md_first field that is used for the pointer to the starting entry in pv_table. Is void * alright, or should I be making some new md_something type for the field?

Aug 24 2020, 11:48 PM
scottph updated the diff for D26132: arm64/pmap: Sparsify pv_table.
Aug 24 2020, 11:45 PM
scottph added a comment to D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.

Instead of modifying vm_phys.c to have some knowledge of superpage reservations, could we instead allocation a reservation array per vm_phys_seg, at least in the VM_PHYSSEG_SPARSE case? I think that would keep the layering clean, and make it easy to solve the issue I pointed out in vm_reserv_from_page().

Aug 24 2020, 11:44 PM
scottph updated the diff for D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 24 2020, 11:42 PM
scottph added inline comments to D26129: Move vm_page_dump bitset array definition to MI code.
Aug 24 2020, 11:37 PM
scottph added inline comments to D26129: Move vm_page_dump bitset array definition to MI code.
Aug 24 2020, 11:35 PM
scottph updated the diff for D26129: Move vm_page_dump bitset array definition to MI code.
Aug 24 2020, 11:32 PM
scottph updated the diff for D26129: Move vm_page_dump bitset array definition to MI code.
Aug 24 2020, 11:30 PM
scottph added a comment to D25798: arm64: Make local stores observable before sending IPIs.

Sorry, I missed the reply somehow.

Don't other PIC drivers require a similar change then? i.e., why isn't the change in pic_ipi_send() instead?

Aug 24 2020, 11:28 PM
scottph updated the diff for D25798: arm64: Make local stores observable before sending IPIs.
Aug 24 2020, 11:27 PM

Aug 21 2020

scottph updated the diff for D26133: arm64/acpi: Give the real PA limit to ACPI.
Aug 21 2020, 7:36 AM
scottph added inline comments to D26129: Move vm_page_dump bitset array definition to MI code.
Aug 21 2020, 7:35 AM
scottph updated the diff for D26129: Move vm_page_dump bitset array definition to MI code.
Aug 21 2020, 7:18 AM
scottph added inline comments to D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 21 2020, 7:18 AM
scottph updated the diff for D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 21 2020, 7:16 AM

Aug 20 2020

scottph added inline comments to D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 20 2020, 9:00 PM
scottph updated the diff for D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 20 2020, 8:56 PM
scottph added a comment to D26134: arm64: Increase dmap size to 95 TiB.

Reviews D26128, D26129, D26130, D26131, D26132, D26133 deal with various arrays that explode in size when the last physical address becomes really huge.

Aug 20 2020, 7:01 AM
scottph requested review of D26134: arm64: Increase dmap size to 95 TiB.
Aug 20 2020, 6:59 AM
scottph requested review of D26133: arm64/acpi: Give the real PA limit to ACPI.
Aug 20 2020, 6:56 AM
scottph requested review of D26132: arm64/pmap: Sparsify pv_table.
Aug 20 2020, 6:55 AM
scottph requested review of D26131: Sparsify the vm_page_dump bitmap.
Aug 20 2020, 6:54 AM
scottph requested review of D26130: vm_reserv: Sparsify the vm_reserv_array when VM_PHYSSEG_SPARSE.
Aug 20 2020, 6:52 AM
scottph added inline comments to D26129: Move vm_page_dump bitset array definition to MI code.
Aug 20 2020, 6:50 AM
scottph requested review of D26129: Move vm_page_dump bitset array definition to MI code.
Aug 20 2020, 6:49 AM
scottph requested review of D26128: bitset: add BIT_FFS_AT() for finding the first bit set greater than a start bit.
Aug 20 2020, 6:48 AM

Jul 28 2020

scottph added inline comments to D25839: efibootmgr: Add option to request booting to the firmware user interface.
Jul 28 2020, 5:06 PM
scottph added inline comments to D25839: efibootmgr: Add option to request booting to the firmware user interface.
Jul 28 2020, 4:29 PM
scottph added a comment to D25839: efibootmgr: Add option to request booting to the firmware user interface.
In D25839#572447, @manu wrote:

LGTM.
I didn't knew this features, looks useful, is the variable reset by the
firmware ? (Or at least is supposed to be ?)

Jul 28 2020, 3:17 PM

Jul 27 2020

scottph requested review of D25839: efibootmgr: Add option to request booting to the firmware user interface.
Jul 27 2020, 11:19 PM

Jul 26 2020

scottph added a comment to D25798: arm64: Make local stores observable before sending IPIs.

smp_rendezvous_cpus() uses a store-release to update smp_rv_waiters before
sending the IPI - why isn't that sufficient?

Jul 26 2020, 11:46 PM

Jul 25 2020

scottph requested review of D25798: arm64: Make local stores observable before sending IPIs.
Jul 25 2020, 4:57 AM

Apr 3 2020

scottph added a comment to D24066: bhyve: support relocating fbuf and passthru data BARs.

It could well be worth investigating if the changes to UEFI to properly support PcdPciDisableBusEnumeration = TRUE are less invasive than allowing runtime BAR changes
(Intel ACRN has this as TRUE)

Apr 3 2020, 6:47 AM · bhyve
scottph updated the diff for D24066: bhyve: support relocating fbuf and passthru data BARs.
Apr 3 2020, 6:43 AM · bhyve

Mar 14 2020

scottph created D24066: bhyve: support relocating fbuf and passthru data BARs.
Mar 14 2020, 3:17 AM · bhyve

Nov 15 2019

scottph added inline comments to D22374: TSX Asynchronous Abort Mitigation.
Nov 15 2019, 8:12 PM

Nov 12 2019

scottph committed rS354648: nvdimm(4): Fix various problems when the using the second label index block.
nvdimm(4): Fix various problems when the using the second label index block
Nov 12 2019, 4:24 PM
scottph closed D22253: nvdimm(4): Fix various problems when the using the second label index block.
Nov 12 2019, 4:24 PM
scottph committed rS354645: nvdimm(4): Only expose namespaces for accessible data SPAs.
nvdimm(4): Only expose namespaces for accessible data SPAs
Nov 12 2019, 3:50 PM
scottph closed D21987: nvdimm(4): Only expose namespaces for accessible data SPAs.
Nov 12 2019, 3:50 PM
scottph updated the diff for D21987: nvdimm(4): Only expose namespaces for accessible data SPAs.
Nov 12 2019, 1:03 AM
scottph updated the diff for D22253: nvdimm(4): Fix various problems when the using the second label index block.

Updated with @cem's suggestions. Reposting for mentor approval.

Nov 12 2019, 12:55 AM

Nov 11 2019

scottph accepted D22007: amd64: Ensure that the state of the switched-out thread is fully flushed.
Nov 11 2019, 6:23 PM

Nov 6 2019

scottph added a comment to D22253: nvdimm(4): Fix various problems when the using the second label index block.
In D22253#486458, @cem wrote:

Thanks!

Some initial tentative feedback and CCing kib@. Caveat, I haven't yet consulted any official documentation nor have I looked at the adjacent code to understand what it is doing, so my comments may just be uninformed. But, provisionally, I don't understand the motivation for some of this changeset.

Nov 6 2019, 7:13 PM
scottph created D22253: nvdimm(4): Fix various problems when the using the second label index block.
Nov 6 2019, 1:34 AM

Nov 4 2019

scottph added inline comments to D22007: amd64: Ensure that the state of the switched-out thread is fully flushed.
Nov 4 2019, 10:48 PM

Nov 1 2019

scottph added a comment to D22007: amd64: Ensure that the state of the switched-out thread is fully flushed.
In D22007#481175, @cem wrote:

Write Combining (WC) ... If the WC buffer is partially filled, the writes may be delayed until the next occurrence of a serializing event; such as, an SFENCE or MFENCE instruction, CPUID execution, a read or write to uncached memory, an interrupt occurrence, or a LOCK instruction execution.

The list is oddly specific and does not include 'mov to control registers,' but the "such as" language and cpuid example suggest to me that setting cr3 may be sufficient (if, as Konstantin points out, we actually did so).

Nov 1 2019, 7:43 PM

Oct 30 2019

scottph committed rP516085: devel/pmdk: update to 1.6.1.
devel/pmdk: update to 1.6.1
Oct 30 2019, 4:21 PM
scottph closed D19504: devel/pmdk: update to 1.6.1.
Oct 30 2019, 4:20 PM

Oct 29 2019

scottph updated the diff for D18028: Add support for Intel Speed Shift.

ah sorry, here's the same patch, but with context restored.

Oct 29 2019, 11:23 PM
scottph updated the diff for D19504: devel/pmdk: update to 1.6.1.

This fixes a failure in the test suite. libvmmalloc was using more thread local storage in the inital-exec mode than rtld was reserving space for, causing the library to fail to load. Switching to the local-dynamic model fixes the issue and allows the test suite to pass.

Oct 29 2019, 11:10 PM

Oct 28 2019

scottph added a comment to D21987: nvdimm(4): Only expose namespaces for accessible data SPAs.
In D21987#480367, @cem wrote:

Do Control Region SPAs have namespaces? I'm still fairly new to NFIT.

Oct 28 2019, 6:14 PM