Paths

Table of Contentst

Remove the i386 kernel and i386-only device drivers
Needs ReviewPublic
Actions

Authored by jhb on Mon, Oct 12, 6:30 PM.

Details

Reviewers

kib
gnn
shurd
brooks
ali_mashtizadeh.com
olce
ngie

Group Reviewers

srcmgr
Restricted Owners Package	(Owns No Changed Paths)
iflib

Summary

This does not move any content from sys/x86 back to sys/amd64. A
few machine headers that are now amd64-specific may make sense to
move in the future.

This also removes all references to DEV_APIC which was only optional
on i386 and previously forced on for amd64. Now the #ifdef's are
removed and just assume DEV_APIC is always present.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 78089
Build 74972: arc lint + arc unit

Event Timeline

jhb created this revision.Mon, Oct 12, 6:30 PM

Owners added a reviewer: Restricted Owners Package.Mon, Oct 12, 6:30 PM

Herald added a reviewer: gnn. · View Herald TranscriptMon, Oct 12, 6:30 PM

Herald added a reviewer: gnn. · View Herald Transcript

Herald added a reviewer: shurd. · View Herald Transcript

Herald added a reviewer: brooks. · View Herald Transcript

Herald added a reviewer: ali_mashtizadeh.com. · View Herald Transcript

Herald added a reviewer: olce. · View Herald Transcript

Herald added a reviewer: ngie. · View Herald Transcript

Herald added a reviewer: iflib. · View Herald Transcript

Herald added subscribers: linuxkpi, quentin.thebault_defenso.fr, olce and 10 others. · View Herald Transcript

jhb requested review of this revision.Mon, Oct 12, 6:30 PM

Harbormaster completed remote builds in B78089: Diff 189649.Mon, Oct 12, 6:30 PM

Herald added a subscriber: obrien. · View Herald TranscriptMon, Oct 12, 6:30 PM

jhb added a parent revision: D60670: universe: More gracefully handle architectures without kernels.Mon, Oct 12, 6:31 PM

jhb added a child revision: D60672: sys: Remove PAE #ifdef's in drivers.

I've split out all the other device driver removals into separate patches, and I've deferred removal of PAE #ifdef's to a followup commit. I could possibly split this up more if there are helpful things to split out, but the actual removal is hard to split further. It may be easier to review this on GitHub by finding the this commit in the list of commits here: https://github.com/freebsd/freebsd-src/compare/main...bsdjhb:freebsd:rm_i386_kernel

I opened a PR that moves i386 to EXTRA_ARCHES a while ago: https://github.com/freebsd/freebsd-src/pull/2052

Should we completely remove i386 from target before this change?

kib added inline comments.Mon, Oct 12, 8:22 PM

sys/cddl/dev/fbt/x86/fbt_isa.c
54	I understand what 0xcc is (INT3) but I am curious what 0xf0 is? From what I remember, it is just a LOCK prefix, how it could be used for patching reliably? This is clearly unrelated to the diff.
sys/crypto/via/padlock.c
35	It is weird to have any such #ifdefs in via/padlock.c, the driver is x86 only.
82	Same.
sys/dev/bxe/bxe_stats.c
33	Does armv7 need this driver, in principle?
sys/dev/dcons/dcons_os.c
320	This is indeed broken on amd64.
sys/dev/virtio/virtqueue.c
569	Again unrelated to the patch, but 1. why use mb() 2. why only on x86?
sys/i386/include/pmap.h
143	Where is this stuff still needed?
sys/i386/include/pmap_pae.h
52	And this?
sys/i386/include/proc.h
37	Are these definitions still needed?
sys/x86/include/apicvar.h
122
124	5 can be reduced, perhaps some time later
sys/x86/pci/pci_bus.c
140
sys/x86/x86/local_apic.c
87	This value is really not a global selector; for amd64 it is ist which was mimicked to gsel to make it similar to i386. Might be worth passing 0 directly and removing the GSEL_APIC symbol altogether.

kbowling added a subscriber: kbowling.Mon, Oct 12, 8:40 PM

kbowling added inline comments.

sys/dev/e1000/e1000_osdep.h
141	This actually highlights the need to fix and use the kern_prefetch KPI but in general this and other drivers are used externally by things like Haiku. The effect here is negligible but I wonder if it's worth the to unifdef stuff like this versus leaving it.

Note (probably to self ;P): we need to gc these from the hw-relnote template (doc/website/archetypes/release/hardware.adoc) after this lands.

In D60671#1391053, @mchoo wrote:

I opened a PR that moves i386 to EXTRA_ARCHES a while ago: https://github.com/freebsd/freebsd-src/pull/2052

Should we completely remove i386 from target before this change?

No, we are keeping the world, and I think we should not move it to EXTRA_ARCHES. i386 world remains Tier 2.

jhb added inline comments.Tue, Oct 13, 3:35 PM

sys/cddl/dev/fbt/x86/fbt_isa.c
54	0xf0 is the old `pop %cs` that became the start of the 2-byte opcode map. That doesn't tell me why you would use it on i386 though. Some later comments imply we use a 2 byte marker on i386. I wonder if it is using UD2 (0x0f 0x0b)?
sys/crypto/via/padlock.c
35	Hmm, true, I might do a separate change to trim these outright. I can maybe do that before the actual removal of i386 bits?
sys/dev/bxe/bxe_stats.c
33	I'm not sure, I suspect it's some PCI card. It should perhaps just be using SIZEOF_LONG instead?
sys/dev/dcons/dcons_os.c
320	I'm not surprised, I think only firewire makes use of dcons(4). Presumably this should just use the dmap?
sys/dev/e1000/e1000_osdep.h
141	Hmmm, I did leave a few #ifdef's in for things I assumed were portable to other OS's. I assumed the `osdep.h` header was OS-specific however (note use of sx_assert a few lines up).
sys/dev/virtio/virtqueue.c
569	Yes, I am not sure why this is here at all. On architectures which need a fence, bus_dmamap_sync() should handle that. virtqueue doesn't use WC memory for queues AFAIK, so the barrier here seems pointless.
sys/i386/include/pmap.h
143	There are some userland bits (I think the zfs stuff in procstat) that need to include vm/pmap.h or the like and that needs some sort of 'struct pmap'. While I could have changed these to be empty, it seemed simpler to just leave them as-is. Note that these are outside of #ifdef _KERNEL. :(
sys/i386/include/pmap_pae.h
52	There is some code in stand/ that needs <machine/pmap_pae.h>. I might not have purged everything possible from this header though.
sys/i386/include/proc.h
37	Hmmm, I thought this mattered for old-school kvm_getprocs() from a crashdump. I could try just making them empty, but code that includes <sys/proc.h> needs `struct mdproc` and `struct mdthread` to still exist.
sys/x86/x86/local_apic.c
87	Ot having some IST_xxx constants?

markj added a subscriber: markj.Tue, Oct 13, 3:52 PM

markj added inline comments.

sys/cddl/dev/fbt/x86/fbt_isa.c
54	The lock prefix is used to cause #UD to be raised. DTrace intercepts both #BP and #UD. I can't remember why both mechanisms exist(ed).

kib added inline comments.Tue, Oct 13, 4:09 PM

sys/cddl/dev/fbt/x86/fbt_isa.c
54	But LOCK only faults if the next instruction is not RMW?
sys/x86/x86/local_apic.c
87	No, we use the numbers directly, in amd64/machdep.c. I see no point.

markj added inline comments.Tue, Oct 13, 4:22 PM

sys/cddl/dev/fbt/x86/fbt_isa.c
54	Yes, I think it is used in a scenario where there is a pre-existing run of nops, created by the kernel linker. To redirect execution, the first nop is overwritten with the lock prefix, and a locked nop will trigger #UD. I still do not understand why they (Solaris) did not instead overwrite the first nop with a breakpoint. That's what we do. Specifically, illumos still uses both tricks: breakpoints for FBT probes (where the first byte of a `push %rbp; mov %rsp, %rbp` prologue is overwritten with int3), and the lock prefix for SDT probes (where a hack is used to create a sequence of nops at each tracepoint during boot). On FreeBSD we use int3 for both types of probes.

Just a couple questions about sys/*crypto. If the headers came from OpenSSL, I think we're ok leaving them be..?

sys/crypto/blake2/config.h
1	Is this from OpenSSL?
sys/crypto/openssl/ossl_aes_gcm.c
1	Is this an OpenSSL provided header, or did we write this using logic from OpenSSL?
sys/crypto/openssl/ossl_sha512.c
2	Is this an OpenSSL provided header, or did we write this using logic from OpenSSL?

jhb added inline comments.Sat, Oct 17, 3:56 PM

sys/x86/x86/local_apic.c
87	Ok, I might do that as a follow-up commit. We don't really need SDT_APIC anymore either.

jhb added inline comments.Sat, Oct 17, 3:58 PM

sys/crypto/blake2/config.h
1	No.
sys/crypto/openssl/ossl_aes_gcm.c
1	The parts being patched are not from OpenSSL and these changes are not relevant to upstream, nor will any upstream changes have any relevance to the parts being changed.
sys/crypto/openssl/ossl_sha512.c
2	Same as above.

jhb marked 17 inline comments as done.Sat, Oct 17, 4:17 PM

jhb added inline comments.

sys/dev/virtio/virtqueue.c
569	Humm, commit 1a92fc9c1210f9c99a19fda5a86682a78d39872f put these back explicitly due to problem reports from users apparently.

jhb added a child revision: D60808: amd64: Close gap between the TLB shootdown and bitmap IPIs.Sat, Oct 17, 4:24 PM

jhb removed a child revision: D60672: sys: Remove PAE #ifdef's in drivers.

Revision Contents
Changeset List

Path

Size

Makefile.inc1

1 line

ObsoleteFiles.inc

55 lines

include/

i386/

Makefile

2 lines

share/

man/

man4/

7 lines

9 lines

9 lines

9 lines

9 lines

man4.i386/

sys/

2 lines

cddl/

dev/

dtrace/

i386/

dtrace_asm.S

dtrace_isa.c

dtrace_subr.c

fbt/

x86/

fbt_isa.c

74 lines

systrace/

systrace.c

7 lines

compat/

freebsd32/

freebsd32.h

2 lines

linprocfs/

linprocfs.c

25 lines

linux/

6 lines

48 lines

2 lines

7 lines

4 lines

4 lines

4 lines

2 lines

34 lines

16 lines

4 lines

22 lines

16 lines

6 lines

25 lines

4 lines

2 lines

linuxkpi/

common/

include/

asm-generic/

io.h

2 lines

asm/

4 lines

2 lines

2 lines

2 lines

2 lines

2 lines

2 lines

2 lines

2 lines

2 lines

linux/

2 lines

4 lines

4 lines

2 lines

4 lines

src/

14 lines

2 lines

2 lines

61 lines

4 lines

x86bios/

x86bios.c

191 lines

conf/

contrib/

dev/

acpica/

acpica_prep.sh

2 lines

crypto/

aesni/

aeskeys_i386.S

aesni.h

2 lines

aesni_ghash.c

17 lines

blake2/

config.h

2 lines

des/

arch/

i386/

openssl/

2 lines

2 lines

2 lines

via/

padlock.c

4 lines

padlock_hash.c

2 lines

dev/

acpica/

28 lines

95 lines

30 lines

16 lines

16 lines

34 lines

2 lines

16 lines

agp/

atkbdc/

6 lines

4 lines

bxe/

bxe_stats.c

11 lines

ciss/

ciss.c

4 lines

cpuctl/

cpuctl.c

14 lines

cxgb/

cxgb_osdep.h

2 lines

cxgb_sge.c

4 lines

cxgbe/

14 lines

5 lines

6 lines

10 lines

dcons/

dcons_crom.c

10 lines

dcons_os.c

18 lines

drm2/

drmP.h

2 lines

drm_os_freebsd.h

7 lines

drm_os_freebsd.c

4 lines

ttm/

ttm_bo_util.c

3 lines

ttm_tt.c

2 lines

e1000/

e1000_osdep.h

2 lines

fb/

fb.c

4 lines

fbreg.h

6 lines

vga.c

11 lines

fdc/

fdc.c

6 lines

hptmv/

osbsd.h

7 lines

hwpmc/

hwpmc_logging.c

4 lines

hwpmc_mod.c

5 lines

hyperv/

pcib/

vmbus_pcib.c

8 lines

vmbus/

i386/

hyperv_machdep.c

vmbus_vector.S

ioat/

ioat_internal.h

5 lines

iommu/

iommu_gas.c

2 lines

ipmi/

ipmi_smbios.c

4 lines

ixgbe/

ixgbe_osdep.h

2 lines

liquidio/

base/

lio_device.h

10 lines

lio_mem_ops.c

28 lines

mlx4/

mlx4_ib/

mlx4_ib_wc.c

4 lines

mvs/

mvs.c

4 lines

ntb/

ntb_hw/

ntb_hw_amd.c

20 lines

ntb_hw_intel.c

20 lines

pcf/

pci/

10 lines

2 lines

4 lines

8 lines

ppc/

ppc.c

44 lines

proto/

proto_core.c

4 lines

qlnx/

qlnxe/

qlnx_def.h

2 lines

qlxgb/

qla_reg.h

2 lines

qlxgbe/

ql_hw.h

2 lines

qlxge/

qls_hw.h

2 lines

random/

unit_test.h

6 lines

sfxge/

common/

efsys.h

6 lines

sfxge_dma.c

2 lines

sge/

if_sge.c

2 lines

sis/

if_sis.c

6 lines

smartpqi/

smartpqi_defines.h

17 lines

smbios/

smbios.c

4 lines

sound/

driver.c

2 lines

pci/

csa.c

27 lines

hda/

hdac.c

6 lines

pcm/

feeder_rate.c

32 lines

spibus/

acpi_spibus.c

8 lines

sym/

sym_hipd.c

2 lines

syscons/

syscons.c

18 lines

tpm/

tpm_bus.c

7 lines

uart/

uart.h

4 lines

uart_dev_ns8250.c

2 lines

virtio/

virtqueue.c

4 lines

vt/

hw/

vga/

vt_vga.c

6 lines

vt_core.c

10 lines

xen/

bus/

xen_intr.c

2 lines

xenpv.c

5 lines

control/

control.c

10 lines

i386/

acpica/

bios/

conf/

i386/

minidump_machdep_base.c

minidump_machdep_nopae.c

minidump_machdep_pae.c

include/

97 lines

326 lines

4 lines

197 lines

48 lines

17 lines

93 lines

38 lines

pc/

86 lines

18 lines

105 lines

27 lines

107 lines

17 lines

77 lines

12 lines

14 lines

91 lines

63 lines

23 lines

68 lines

25 lines

17 lines

xen/

linux/

linux_dummy_machdep.c

linux_ptrace_machdep.c

linux_syscall.h

linux_syscalls.c

linux_sysent.c

linux_systrace_args.c

pci/

isa/

4 lines

3 lines

kern/

38 lines

4 lines

2 lines

4 lines

2 lines

27 lines

44 lines

4 lines

10 lines

4 lines

2 lines

17 lines

100 lines

4 lines

4 lines

libkern/

gsb_crc32.c

4 lines

modules/

Makefile

24 lines

agp/

Makefile

9 lines

bios/

Makefile

4 lines

smapi/

Makefile

cpufreq/

Makefile

6 lines

dtrace/

Makefile

4 lines

dtrace/

Makefile

6 lines

dtraceall/

dtraceall.c

4 lines

fasttrap/

Makefile

2 lines

fbt/

Makefile

2 lines

kinst/

Makefile

2 lines

hyperv/

vmbus/

Makefile

3 lines

i2c/

controllers/

pcf/

Makefile

9 lines

linux/

Makefile

35 lines

linuxkpi/

Makefile

4 lines

ossl/

Makefile

9 lines

zfs/

zfs_config.h

4 lines

net/

altq/

14 lines

9 lines

2 lines

2 lines

2 lines

2 lines

netgraph/

ng_source.c

4 lines

opencrypto/

crypto.c

4 lines

sys/

2 lines

2 lines

3 lines

2 lines

2 lines

6 lines

vm/

2 lines

4 lines

15 lines

4 lines

x86/

cpufreq/

include/

4 lines

24 lines

2 lines

18 lines

70 lines

5 lines

17 lines

4 lines

iommu/

17 lines

56 lines

24 lines

14 lines

isa/

atpic.c

5 lines

clock.c

11 lines

linux/

linux_x86_sigframe.h

4 lines

pci/

pci_bus.c

283 lines

x86/

33 lines

21 lines

20 lines

4 lines

872 lines

17 lines

29 lines

29 lines

18 lines

15 lines

6 lines

13 lines

37 lines

41 lines

2 lines

xen/

hvm.c

6 lines

xen_apic.c

50 lines

Remove the i386 kernel and i386-only device driversNeeds ReviewPublicActions

Details

Diff Detail

Event Timeline

Revision ContentsChangeset List

Diff 189649

Makefile.inc1

ObsoleteFiles.inc

include/i386/Makefile

share/man/man4/cpufreq.4

share/man/man4/hpt27xx.4

share/man/man4/hptmv.4

share/man/man4/hptnr.4

share/man/man4/hptrr.4

share/man/man4/man4.i386/CPU_ELAN.4

share/man/man4/man4.i386/Makefile

share/man/man4/man4.i386/Makefile.depend

share/man/man4/man4.i386/longrun.4

share/man/man4/man4.i386/npx.4

share/man/man4/man4.i386/pae.4

share/man/man4/man4.i386/pnp.4

share/man/man4/man4.i386/pnpbios.4

share/man/man4/man4.i386/smapi.4

sys/Makefile

sys/cddl/dev/dtrace/i386/dtrace_asm.S

sys/cddl/dev/dtrace/i386/dtrace_isa.c

sys/cddl/dev/dtrace/i386/dtrace_subr.c

sys/cddl/dev/fbt/x86/fbt_isa.c

sys/cddl/dev/systrace/systrace.c

sys/compat/freebsd32/freebsd32.h

sys/compat/linprocfs/linprocfs.c

sys/compat/linux/linux_event.c

sys/compat/linux/linux_file.c

sys/compat/linux/linux_futex.c

sys/compat/linux/linux_ioctl.c

sys/compat/linux/linux_ipc.h

sys/compat/linux/linux_ipc.c

sys/compat/linux/linux_ipc64.h

sys/compat/linux/linux_misc.h

sys/compat/linux/linux_misc.c

sys/compat/linux/linux_signal.c

sys/compat/linux/linux_socket.h

sys/compat/linux/linux_socket.c

sys/compat/linux/linux_stats.c

sys/compat/linux/linux_time.h

sys/compat/linux/linux_time.c

sys/compat/linux/linux_timer.c

sys/compat/linux/linux_vdso.c

sys/compat/linuxkpi/common/include/asm-generic/io.h

sys/compat/linuxkpi/common/include/asm/atomic.h

sys/compat/linuxkpi/common/include/asm/barrier.h

sys/compat/linuxkpi/common/include/asm/cpu_device_id.h

sys/compat/linuxkpi/common/include/asm/cpufeature.h

sys/compat/linuxkpi/common/include/asm/hypervisor.h

sys/compat/linuxkpi/common/include/asm/memtype.h

sys/compat/linuxkpi/common/include/asm/pgtable.h

sys/compat/linuxkpi/common/include/asm/processor.h

sys/compat/linuxkpi/common/include/asm/smp.h

sys/compat/linuxkpi/common/include/asm/topology.h

sys/compat/linuxkpi/common/include/linux/acpi.h

sys/compat/linuxkpi/common/include/linux/io.h

sys/compat/linuxkpi/common/include/linux/kconfig.h

sys/compat/linuxkpi/common/include/linux/kernel.h

sys/compat/linuxkpi/common/include/linux/page.h

sys/compat/linuxkpi/common/src/linux_compat.c

sys/compat/linuxkpi/common/src/linux_current.c

sys/compat/linuxkpi/common/src/linux_fpu.c

sys/compat/linuxkpi/common/src/linux_page.c

sys/compat/linuxkpi/common/src/linux_pci.c

sys/compat/x86bios/x86bios.c

sys/conf/Makefile.i386

sys/conf/files.i386

sys/conf/ldscript.i386

sys/conf/ldscript.kmod.i386

sys/conf/options.i386

sys/contrib/dev/acpica/acpica_prep.sh

sys/crypto/aesni/aeskeys_i386.S

sys/crypto/aesni/aesni.h

sys/crypto/aesni/aesni_ghash.c

sys/crypto/blake2/config.h

Remove the i386 kernel and i386-only device drivers
Needs ReviewPublic
Actions

Revision Contents
Changeset List