Page MenuHomeFreeBSD
Feed Advanced Search

Oct 3 2019

kib committed rS353033: Remove unnecessary vm/vm_page.h and vm/vm_pager.h includes from.
Remove unnecessary vm/vm_page.h and vm/vm_pager.h includes from
Oct 3 2019, 8:25 AM
kib closed D21881: Remove unnecessary vm_page.h and vm_pager.h includes..
Oct 3 2019, 8:25 AM
kib accepted D21881: Remove unnecessary vm_page.h and vm_pager.h includes..
Oct 3 2019, 8:24 AM
kib created D21883: Fix interface between nfsclient and vnode pager..
Oct 3 2019, 8:18 AM

Oct 2 2019

kib accepted D21859: Tighten the default permissions on preloaded files..
Oct 2 2019, 7:33 PM
kib accepted D21860: Set permissions appropriately on preloaded DSOs..
Oct 2 2019, 7:15 PM
kib added inline comments to D21859: Tighten the default permissions on preloaded files..
Oct 2 2019, 7:12 PM
kib accepted D21876: simplify path handling in sysctl_try_reclaim_vnode.
Oct 2 2019, 5:10 PM
kib added inline comments to D21876: simplify path handling in sysctl_try_reclaim_vnode.
Oct 2 2019, 4:44 PM
kib added inline comments to D21876: simplify path handling in sysctl_try_reclaim_vnode.
Oct 2 2019, 2:56 PM
kib accepted D21871: nvdimm: Fix error path mis-free.
Oct 2 2019, 2:14 PM
kib added inline comments to D21876: simplify path handling in sysctl_try_reclaim_vnode.
Oct 2 2019, 2:09 PM
kib committed rS353007: MFC r352684:.
MFC r352684:
Oct 2 2019, 1:46 PM
kib committed rS353006: MFC r352684:.
MFC r352684:
Oct 2 2019, 1:43 PM
kib accepted D21862: Use OBJT_PHYS objects for linker files..
Oct 2 2019, 1:40 PM

Oct 1 2019

kib added a comment to D21862: Use OBJT_PHYS objects for linker files..

No objections, but why ? link_elf_obj also uses OBJT_DEFAULT.

Oct 1 2019, 9:21 PM
kib accepted D21864: Disallow fcntl(F_READAHEAD) when the vnode is not a regular file..
Oct 1 2019, 9:14 PM
kib added a comment to D21845: linuxolator: implement memfd_create syscall.

I think you need to provide some explanation how your l/b bits functions are supposed to work.

Oct 1 2019, 4:53 PM
kib accepted D21822: Provide generic sub-word atomic *cmpset.
Oct 1 2019, 2:50 PM

Sep 30 2019

kib accepted D21836: mips: fcmpset: do not spin on sc failure.
Sep 30 2019, 9:07 PM
kib accepted D21852: linux_renameat2: don't add extra \n on error.
Sep 30 2019, 4:25 PM
kib added inline comments to D21822: Provide generic sub-word atomic *cmpset.
Sep 30 2019, 2:29 PM
kib added inline comments to D21836: mips: fcmpset: do not spin on sc failure.
Sep 30 2019, 2:25 PM
kib added a comment to D21845: linuxolator: implement memfd_create syscall.

I dislike the reliance on equal FreeBSD and Linux flags values. Please use LINUX_MFD_XXX constants in linuxolator. You may add _Static_asserts that they are equal and avoid translating the current flags, but I do not think that it is sustainable in long run.

Sep 30 2019, 2:21 PM
kib committed rS352900: MFC r352620, MFS12 r352892:.
MFC r352620, MFS12 r352892:
Sep 30 2019, 2:17 PM
kib committed rS352893: MFC r352620:.
MFC r352620:
Sep 30 2019, 7:43 AM
kib committed rS352892: MFC r352620:.
MFC r352620:
Sep 30 2019, 7:42 AM

Sep 29 2019

kib accepted D21823: Remove page locking from pmap_mincore()..
Sep 29 2019, 8:08 PM
kib added a comment to D20674: [libkvm] kvm_geswapinfo() to stop subtracting vm.dmmax size from swap device sizes..
In D20674#476867, @alc wrote:
Sep 29 2019, 8:05 PM
kib added inline comments to D21832: amd64 pmap: batch chunk removal in pmap_remove_pages.
Sep 29 2019, 7:55 PM
kib added a comment to D20674: [libkvm] kvm_geswapinfo() to stop subtracting vm.dmmax size from swap device sizes..
In D20674#476856, @alc wrote:
In D20674#476783, @kib wrote:

That said, I suspect that we no longer need to avoid initial blocks. At least not for the label preservation.

I'm afraid that is not true. I found out the hard way that it is needed. I lost a disk label, swap area, and file system when I first enabled "trimonce". However, the configuration of that storage device is usual. The first partition is a swap area, not a file system.

Sep 29 2019, 4:32 PM
kib accepted D21836: mips: fcmpset: do not spin on sc failure.
Sep 29 2019, 4:29 PM
kib added inline comments to D21836: mips: fcmpset: do not spin on sc failure.
Sep 29 2019, 3:19 PM
kib added inline comments to D21833: amd64 pmap: retire the global lock array in favor of per-superpage locks.
Sep 29 2019, 2:49 PM
kib added inline comments to D21822: Provide generic sub-word atomic *cmpset.
Sep 29 2019, 2:35 PM
kib added a comment to D21823: Remove page locking from pmap_mincore()..

I think the approach of relying on the object lock is fine.

Sep 29 2019, 2:18 PM
kib added inline comments to D21822: Provide generic sub-word atomic *cmpset.
Sep 29 2019, 2:11 PM
kib added a comment to D20674: [libkvm] kvm_geswapinfo() to stop subtracting vm.dmmax size from swap device sizes..
In D20674#476657, @alc wrote:

I believe that there is a sound reason for not reporting the unmodified size of the device. Specifically, the swap pager can't use the first, if I recall correctly, 8KB of a swap device because it might contain a disk label. However, I can't explain why we use dmmax to calculate the modified size.

Sep 29 2019, 2:06 PM
kib added inline comments to D21822: Provide generic sub-word atomic *cmpset.
Sep 29 2019, 10:16 AM
kib accepted D21832: amd64 pmap: batch chunk removal in pmap_remove_pages.
Sep 29 2019, 9:31 AM

Sep 28 2019

kib committed rS352833: MFC r352582, r352585:.
MFC r352582, r352585:
Sep 28 2019, 7:33 AM

Sep 27 2019

kib closed D21566: Improve MD page fault handlers..
Sep 27 2019, 6:44 PM
kib committed rS352807: Improve MD page fault handlers..
Improve MD page fault handlers.
Sep 27 2019, 6:43 PM
kib accepted D20674: [libkvm] kvm_geswapinfo() to stop subtracting vm.dmmax size from swap device sizes..
Sep 27 2019, 1:54 PM
kib added a comment to D20575: pseudofs: add sysctl to set sbuf buffer size.

Why do you need a sysctl ? If bumping, simply increase to some larger but reasonable value, e.g. 2M. IOSIZE_MAX is definitely unreasonable.

Sep 27 2019, 1:53 PM
kib accepted D21813: nvdimm(4): Extract ACPI root bus driver.
Sep 27 2019, 7:18 AM
kib added a comment to D21813: nvdimm(4): Extract ACPI root bus driver.

Is there any public specs about NVDIMM enumeration using e820 ?

Sep 27 2019, 7:18 AM
kib updated the diff for D21566: Improve MD page fault handlers..

More notes by alc.

Sep 27 2019, 7:15 AM

Sep 26 2019

kib accepted D21596: (vm object 6) Convert pmap_is_modified, pmap_remove_write, and pmap_clear_modify to expecting busy to be held by the caller now that all callers hold it..
Sep 26 2019, 8:59 PM
kib accepted D21594: (vm object 4) Wrap valid accesses in inlines. Add missing busy acquires to protect valid..
Sep 26 2019, 8:57 PM
kib accepted D21791: Fix a race in vm_page_swapqueue()..
Sep 26 2019, 8:46 PM
kib accepted D21810: Add myself (kaktus) as a src commiter.
Sep 26 2019, 8:42 PM
kib accepted D21811: Add myself (kaktus) as a src commiter.
Sep 26 2019, 8:41 PM
kib updated the diff for D21566: Improve MD page fault handlers..

More alc' notes.

Sep 26 2019, 8:27 PM
kib added inline comments to D21566: Improve MD page fault handlers..
Sep 26 2019, 8:23 PM
kib added inline comments to D21811: Add myself (kaktus) as a src commiter.
Sep 26 2019, 8:19 PM
kib accepted D21788: Document the convention of adding a blank line before comments..
Sep 26 2019, 5:15 PM
kib added inline comments to D21740: lockprof: fix hangs under load when changing the state or dumping stats.
Sep 26 2019, 7:35 AM
kib added a comment to D21646: vfs: add root vnode caching for mount points.

I stil has an opinion that the mp root vnode caching should be better integrated with existing code. Right now it feels as a side hack.

Sep 26 2019, 7:28 AM
kib added a comment to D21703: o Unify all <machine>/resource.h, unhide PCI_RES_BUS, add CLK and PWR..
In D21703#475705, @ray wrote:

Return back #ifdef NEW_PCIB, @jhb not sure it is safe to omit.

Sep 26 2019, 6:01 AM

Sep 25 2019

kib accepted D21790: Fix object unlocking in vm_object_unwire() after r352174..
Sep 25 2019, 6:44 PM
kib added a comment to D21788: Document the convention of adding a blank line before comments..

I was told by bde (?) that after the block of code that is described by the multi-line comment, there should be a blank line. Since you carefully document other moments, this one should be described as well, I believe.

Sep 25 2019, 6:43 PM
kib accepted D21785: remove obsolete i386 MD memchr implementation.
Sep 25 2019, 2:35 PM
kib closed D21777: x86: Fall back to leaf 0x16 if TSC frequency is obtained by CPUID and leaf 0x15 is not functional..
Sep 25 2019, 1:37 PM
kib committed rS352684: x86: Fall back to leaf 0x16 if TSC frequency is obtained by CPUID and.
x86: Fall back to leaf 0x16 if TSC frequency is obtained by CPUID and
Sep 25 2019, 1:37 PM
kib committed rS352683: MFC r352488:.
MFC r352488:
Sep 25 2019, 1:30 PM
kib committed rS352682: MFC r352486:.
MFC r352486:
Sep 25 2019, 1:28 PM
kib accepted D21740: lockprof: fix hangs under load when changing the state or dumping stats.
Sep 25 2019, 12:40 PM
kib added a comment to D21566: Improve MD page fault handlers..

Alan, any more comments ? I want to close the review.

Sep 25 2019, 12:35 PM
kib added a comment to D21767: Fix handling of page wirings exec_map_first_page()..

Here is a program that reproduces the panic: https://reviews.freebsd.org/P319
The program that syzkaller generated uses fewer system calls but is multi-threaded.

Sep 25 2019, 12:33 PM
kib added inline comments to D21762: Add a sleep lock to the NFS node to protect n_size.
Sep 25 2019, 12:25 PM
kib accepted D21393: [3/3] Add linux-compatible memfd_create.
Sep 25 2019, 12:20 PM

Sep 24 2019

kib added a comment to D21767: Fix handling of page wirings exec_map_first_page()..
In D21767#475356, @alc wrote:

Do we actually allow truncation on a running executable file?

Sep 24 2019, 6:35 PM
kib added inline comments to D21783: Assert that the PGA_{EXECUTABLE,WRITEABLE} flags do not leak..
Sep 24 2019, 5:21 PM
kib accepted D21782: Add some counters for per-page queue operations..
Sep 24 2019, 5:20 PM
kib added inline comments to D21783: Assert that the PGA_{EXECUTABLE,WRITEABLE} flags do not leak..
Sep 24 2019, 5:14 PM
kib accepted D21783: Assert that the PGA_{EXECUTABLE,WRITEABLE} flags do not leak..

Would pmap_page_is_mapped() also a good or even better place to put such asserts ?

Sep 24 2019, 5:12 PM
kib added a comment to D21775: (WIP) Rewrite makesyscalls.sh in Lua.

Note a use case which must not be broken: often the workstation where syscall files are regenerated is running some kind of stable or even a release.

Sep 24 2019, 2:37 PM
kib created D21777: x86: Fall back to leaf 0x16 if TSC frequency is obtained by CPUID and leaf 0x15 is not functional..
Sep 24 2019, 10:22 AM
kib added a comment to D21774: [wip] vfs: switch vnode count management away from refcounts to atomics.

Perhaps name it atomic_ops.h or atomic_algo.h.

Sep 24 2019, 9:53 AM
kib added inline comments to D21740: lockprof: fix hangs under load when changing the state or dumping stats.
Sep 24 2019, 9:48 AM
kib committed rS352650: MFC r352459:.
MFC r352459:
Sep 24 2019, 6:37 AM
kib committed rS352649: MFC r352458:.
MFC r352458:
Sep 24 2019, 6:36 AM
kib committed rS352648: MFC r352459:.
MFC r352459:
Sep 24 2019, 6:32 AM
kib committed rS352647: MFC r352458:.
MFC r352458:
Sep 24 2019, 6:30 AM
kib committed rS352646: MFC r352456:.
MFC r352456:
Sep 24 2019, 6:23 AM
kib committed rS352645: MFC r352455:.
MFC r352455:
Sep 24 2019, 6:21 AM
kib committed rS352644: MFC r352456:.
MFC r352456:
Sep 24 2019, 6:20 AM
kib committed rS352643: MFC r352455:.
MFC r352455:
Sep 24 2019, 6:19 AM

Sep 23 2019

kib accepted D21767: Fix handling of page wirings exec_map_first_page()..
Sep 23 2019, 6:39 PM
kib accepted D21767: Fix handling of page wirings exec_map_first_page()..
Sep 23 2019, 6:22 PM
kib added a comment to D21767: Fix handling of page wirings exec_map_first_page()..
In D21767#474999, @kib wrote:

What is the point of deactivating them ? Inactive scan frees invalid pages upon encountering them.

As opposed to doing nothing? vm_page_deactivate_noreuse() ensures that the pages skip LRU, so the page daemon (which handles races with concurrent wirings) will free them sooner.

Sep 23 2019, 6:10 PM
kib accepted D21768: Remove the wire_count field from struct vm_page..
Sep 23 2019, 6:08 PM
kib added a comment to D21767: Fix handling of page wirings exec_map_first_page()..

What is the point of deactivating them ? Inactive scan frees invalid pages upon encountering them.

Sep 23 2019, 6:05 PM
kib added a comment to D21762: Add a sleep lock to the NFS node to protect n_size.

So, do you think that just replacing n_mtx with n_solock makes more sense?

Yes.

Sep 23 2019, 5:16 PM
kib accepted D21758: Introduce pmap_change_prot()..
Sep 23 2019, 5:13 PM
kib accepted D21756: Fix some problems in SPARSE_MAPPING in the kernel linker..
Sep 23 2019, 5:11 PM
kib added a comment to D21756: Fix some problems in SPARSE_MAPPING in the kernel linker..

Is there any way to enable SPARSE_MAPPING except manually adding #define ?

Sep 23 2019, 2:49 PM
kib added inline comments to D21758: Introduce pmap_change_prot()..
Sep 23 2019, 2:39 PM
kib committed rS352621: MFC r352393, r352457:.
MFC r352393, r352457:
Sep 23 2019, 1:46 PM