Page MenuHomeFreeBSD

dougm (Doug Moore)
User

Projects

User does not belong to any projects.

User Details

User Since
Jun 30 2017, 3:18 PM (103 w, 4 d)

Tauist. Bit twiddler. Advocate for continued logarithms.

Recent Activity

Today

dougm added a comment to D20664: Changes to vm_map_lookup_entry.

It broke boot, so I've reverted it. Now, while I look for the problem, I'll accept your shaming.

Wed, Jun 26, 3:17 AM
dougm committed rS349405: Revert r349393, which leads to an assertion failure on bootup, in….
Revert r349393, which leads to an assertion failure on bootup, in…
Wed, Jun 26, 3:13 AM

Yesterday

dougm added a comment to D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE.

Fine-tune a comment that still includes "now will do cow".

Tue, Jun 25, 10:55 PM
D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE now requires review to proceed.

Discard the parts of a comment that may mislead.

Tue, Jun 25, 8:34 PM
dougm committed rS349393: Eliminate some uses of the prev and next fields of vm_map_entry_t..
Eliminate some uses of the prev and next fields of vm_map_entry_t.
Tue, Jun 25, 8:25 PM
dougm closed D20664: Changes to vm_map_lookup_entry.
Tue, Jun 25, 8:25 PM
dougm added a comment to D20664: Changes to vm_map_lookup_entry.

Apply reviewer suggestions.

Tue, Jun 25, 7:28 PM
dougm committed rS349379: Document the KERN_PROTECTION_FAILURE return value from vm_map_protect()..
Document the KERN_PROTECTION_FAILURE return value from vm_map_protect().
Tue, Jun 25, 5:27 PM
Herald added a reviewer for D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value.: manpages.
Tue, Jun 25, 5:27 PM
D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE now requires review to proceed.

Be more articulate.

Tue, Jun 25, 5:09 PM
dougm accepted D20758: Add a return value to vm_page_remove()..
Tue, Jun 25, 5:03 PM
D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE now requires review to proceed.

Back-up one version.

Tue, Jun 25, 5:02 PM
dougm added a comment to D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..

Articulate.

Tue, Jun 25, 4:51 PM
D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value. now requires review to proceed.

Accept reviewer suggestion.

Tue, Jun 25, 4:47 PM
dougm added a comment to D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE.

Don't over-simplify.

Tue, Jun 25, 4:45 PM
dougm added inline comments to D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE.
Tue, Jun 25, 4:42 PM
dougm added inline comments to D20758: Add a return value to vm_page_remove()..
Tue, Jun 25, 4:32 PM
dougm added a comment to D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE.

Accept reviewer advice.

Tue, Jun 25, 3:51 PM
dougm updated the diff for D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..
Tue, Jun 25, 3:40 PM
dougm added a comment to D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..

Accept Kostik's suggestion.

Tue, Jun 25, 3:37 PM
dougm created D20753: simplify entries after vm_map_protect -> KERN_RESOURCE_SHORTAGE.
Tue, Jun 25, 8:38 AM
dougm added a comment to D20664: Changes to vm_map_lookup_entry.

Resolve a conflict.

Tue, Jun 25, 8:12 AM
dougm closed D20711: Stop early clipping in vm_map_protect.
Tue, Jun 25, 7:45 AM
dougm committed rS349363: vm_map_protect may return an INVALID_ARGUMENT or PROTECTION_FAILURE.
vm_map_protect may return an INVALID_ARGUMENT or PROTECTION_FAILURE
Tue, Jun 25, 7:44 AM
dougm added a comment to D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..

Take Alan's suggestion.

Tue, Jun 25, 7:24 AM
dougm updated subscribers of D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..
Tue, Jun 25, 6:47 AM
dougm created D20751: vm_map_protect(9) fails to document KERN_RESOURCE_SHORTAGE return value..
Tue, Jun 25, 6:31 AM

Mon, Jun 24

dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Remove redundant test.

Mon, Jun 24, 7:40 PM
dougm updated the summary of D20711: Stop early clipping in vm_map_protect.
Mon, Jun 24, 7:46 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Retreat. I know I can improve something, and I know that I'm misunderstanding something else. I'll try to take the small win for now.

Mon, Jun 24, 5:29 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Move the simplification of entries to a separate, third loop, not because I want to but because I want to see if simplification is what's making my accounting assertion fail, or if the cause is some other thing I don't understand.

Mon, Jun 24, 4:33 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Add an assertion that the amount of swap space to reserved has been calculated correctly.

Mon, Jun 24, 2:35 AM

Sun, Jun 23

dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Remove a redundant condition, one which is tested for at the top of the loop.

Sun, Jun 23, 11:17 PM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

If vm_map_entry_back is going to be invoked by clipping, then use the size of the untrimmed entry, which is what vm_map_entry_back will set as the size of the object that it creates. Otherwise, if there's no object, there won't be an object made and use the clipped map entry difference. Otherwise, use the size from the pre-existing object, the size of which won't be modified by vm_map_entry_charge_object, or by clip_start or clip_end.

Sun, Jun 23, 11:14 PM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Don't assume I can ignore ptoa(obj->size). Account for clipping when there's no object.

Sun, Jun 23, 8:56 PM
dougm added inline comments to D20711: Stop early clipping in vm_map_protect.
Sun, Jun 23, 8:38 PM
dougm added inline comments to D20711: Stop early clipping in vm_map_protect.
Sun, Jun 23, 8:31 PM
dougm added inline comments to D20711: Stop early clipping in vm_map_protect.
Sun, Jun 23, 7:51 PM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Address easy suggestions from reviewer.

Sun, Jun 23, 7:36 PM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Don't account for clipping in computing the amount of swap space to reserve. Fix some style glitches.

Sun, Jun 23, 8:32 AM

Sat, Jun 22

dougm committed rS349293: You can't use block special nodes for swap, so don't let that happen..
You can't use block special nodes for swap, so don't let that happen.
Sat, Jun 22, 7:27 PM
dougm closed D20723: Remove bad /useless S_ISBLK from swapon.
Sat, Jun 22, 7:27 PM
D20723: Remove bad /useless S_ISBLK from swapon now requires review to proceed.

I understand that I violated a style(9) header rule too. Fix that.

Sat, Jun 22, 5:37 PM
dougm created D20723: Remove bad /useless S_ISBLK from swapon.
Sat, Jun 22, 5:16 PM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Reorder events in the final loop now that the significance of 'cred' has been partially revealed to me.

Sat, Jun 22, 7:56 AM
dougm added a comment to D20633: Does clip_start require entry simplification?.
In D20633#448011, @pho wrote:

I'm almost done with the amd64 tests and only found one instance of the printf:

20190621 13:21:33 all (235/646): mmap14.sh
_vm_map_clip_start: simplifying entry start 7fffdfdfe000 end 7fffdffde000 next_read 7fffdfdfd000 max_free 7ff7de9fc000 eflags 30000 object-type -1

Shall I move on to D20711?

Sat, Jun 22, 6:18 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Don't allocate swap space until the in-transition test is passed.

Sat, Jun 22, 5:44 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Relative to the previous version, combine the last two loops of vm_map_protect into one.

Sat, Jun 22, 5:40 AM
dougm added a comment to D20711: Stop early clipping in vm_map_protect.

Reduce the swap_reserve calls to one, and not clip or modify map entry fields until it succeeds. This change makes an assumption about the need to check ptoa(obj->size) - that it isn't necessary when you the start and end of the map entry - that I'm believing because alc told me so.

Sat, Jun 22, 4:42 AM
dougm retitled D20711: Stop early clipping in vm_map_protect from Stop early clipping in vm_map_reserve to Stop early clipping in vm_map_protect.
Sat, Jun 22, 3:55 AM
dougm committed rS349286: Modify swapon(8) to invoke BIO_DELETE to trim swap devices, either if.
Modify swapon(8) to invoke BIO_DELETE to trim swap devices, either if
Sat, Jun 22, 3:16 AM
Herald added a reviewer for D20599: simple trimming before swapon: manpages.
Sat, Jun 22, 3:16 AM

Fri, Jun 21

dougm added inline comments to D20664: Changes to vm_map_lookup_entry.
Fri, Jun 21, 10:47 PM
dougm added a comment to D20664: Changes to vm_map_lookup_entry.

Apply most reviewer suggestions. Explain the one I didn't apply.

Fri, Jun 21, 10:47 PM
dougm added a comment to D20579: Remove limit on blist_alloc maxsize, allow bigger getswapspace allocations.

By renaming, re-indenting, and updating stale comments, seek to answer questions and address reviewer concerns.

Fri, Jun 21, 10:20 PM
dougm added a comment to D20599: simple trimming before swapon.

Add a header. Drop an extra argument to warn().

Fri, Jun 21, 6:34 PM
dougm added inline comments to D20599: simple trimming before swapon.
Fri, Jun 21, 3:37 AM
dougm added inline comments to D20711: Stop early clipping in vm_map_protect.
Fri, Jun 21, 3:12 AM

Thu, Jun 20

dougm added a comment to D20633: Does clip_start require entry simplification?.

Just to be clear, I understand that this latest patch produces no printfs on amd64, but only on i386, and the previous version produces printfs on both architectures. Is that understanding correct?

Thu, Jun 20, 4:26 PM
dougm created D20711: Stop early clipping in vm_map_protect.
Thu, Jun 20, 4:19 PM
dougm added a comment to D20633: Does clip_start require entry simplification?.
In D20633#447540, @pho wrote:

I see lots of _vm_map_clip_start on i386 with D20633.58825.diff.
https://people.freebsd.org/~pho/stress/log/dougm034.txt

Thu, Jun 20, 3:06 PM
dougm added a comment to D20633: Does clip_start require entry simplification?.

I've identified a case where vm_map_protect leaves an entry un-simplified, and attempted to address it. I hope to discover whether the number of simplify_entry calls logged as necessary in clip_start is reduced by this change.

Thu, Jun 20, 6:02 AM

Wed, Jun 19

dougm added a comment to D20599: simple trimming before swapon.

Addressed reviewer comments about these changes.

Wed, Jun 19, 8:14 PM

Tue, Jun 18

dougm added a comment to D20664: Changes to vm_map_lookup_entry.

... but keep the boolean_t to bool return value transition for vm_map_entry_lookup.

Tue, Jun 18, 4:00 AM
dougm updated the summary of D20664: Changes to vm_map_lookup_entry.

Too many places depend on the vm_map_entry_lookup name, so leave it alone and rename something else.

Tue, Jun 18, 3:56 AM
dougm added a comment to D20675: Eliminate incorrect 2 pagesized block skipping while preparing swap device..

Most of these changes are distractions, unrelated to your real purpose. I suggest you get rid of them. What's left is you removing a bug fix from 2003 without any explanation about why that fix is no longer necessary, and you fixing a bug that hasn't been filed with no real explanation of how it is a fix. I don't think anyone has enough information to approve this change.

Tue, Jun 18, 3:27 AM

Mon, Jun 17

dougm added inline comments to D13484: swapoff to release multiple swap blocks at a time.
Mon, Jun 17, 10:21 PM
dougm added a comment to D20599: simple trimming before swapon.

Endeavor to apply reviewer suggestions.

Mon, Jun 17, 10:14 PM
dougm added a reviewer for D20674: [libkvm] kvm_geswapinfo() to stop subtracting vm.dmmax size from swap device sizes.: alc.
Mon, Jun 17, 8:50 PM
dougm added a reviewer for D20675: Eliminate incorrect 2 pagesized block skipping while preparing swap device.: alc.
Mon, Jun 17, 8:48 PM

Sun, Jun 16

dougm created D20664: Changes to vm_map_lookup_entry.
Sun, Jun 16, 4:39 PM

Sat, Jun 15

dougm added a comment to D20633: Does clip_start require entry simplification?.

Just to be clear, I understand you to be saying that the map11 test produces the four lines of output above and no other tests, including buildworld/installworld, produce any such output. Is that correct?

Sat, Jun 15, 4:05 PM
dougm committed rS349046: Critical comments were lost in r349203. This patch seeks to restore.
Critical comments were lost in r349203. This patch seeks to restore
Sat, Jun 15, 4:30 AM
dougm closed D20632: Restore commentary lost in r349203.
Sat, Jun 15, 4:30 AM
dougm updated the diff for D20635: release multiple swap blocks at a time.
Sat, Jun 15, 4:18 AM
dougm retitled D20635: release multiple swap blocks at a time from The patch after merging in small helper functions to release multiple swap blocks at a time.
Sat, Jun 15, 4:18 AM

Fri, Jun 14

dougm planned changes to D20561: Avoid locking, manipulating vm_map for bogus address lookup.
Fri, Jun 14, 9:36 PM
dougm updated the diff for D20633: Does clip_start require entry simplification?.

Make it compile.

Fri, Jun 14, 6:17 PM
dougm updated the diff for D20632: Restore commentary lost in r349203.

Apply reviewer suggestion.

Fri, Jun 14, 4:31 PM
dougm updated the diff for D20633: Does clip_start require entry simplification?.

Write information about the cases in which clip_start requires entry simplification, with the hope that the origins of those cases can be identified and that simplification happen elsewhere.

Fri, Jun 14, 4:10 PM
dougm created D20635: release multiple swap blocks at a time.
Fri, Jun 14, 8:05 AM
dougm created D20633: Does clip_start require entry simplification?.
Fri, Jun 14, 6:18 AM
dougm updated the diff for D20632: Restore commentary lost in r349203.

Leave out unrelated patches.

Fri, Jun 14, 3:53 AM
dougm updated the diff for D20632: Restore commentary lost in r349203.

Separate the comments and code blocks for 'create backing object' from 'clone entry'.

Fri, Jun 14, 3:42 AM
dougm committed rS349030: Avoid using the prev field of vm_map_entry_t in two functions that.
Avoid using the prev field of vm_map_entry_t in two functions that
Fri, Jun 14, 3:16 AM
dougm closed D20628: Use vm_map_entry_t field prev less.
Fri, Jun 14, 3:16 AM
dougm added a comment to D13484: swapoff to release multiple swap blocks at a time.

Can you update this patch, please, at least so that it can be applied? Part of it was put into a separate patch and has already been put in head.

Fri, Jun 14, 2:29 AM
dougm created D20632: Restore commentary lost in r349203.
Fri, Jun 14, 2:04 AM
dougm updated the diff for D20625: clarify reserv_test_config.

Update comments in accordance with suggestions from alc.

Fri, Jun 14, 1:40 AM

Thu, Jun 13

dougm updated the diff for D20599: simple trimming before swapon.

Address reviewer comments.

Thu, Jun 13, 8:23 PM
dougm committed rS349023: Create a function for creating objects to back map entries, and one.
Create a function for creating objects to back map entries, and one
Thu, Jun 13, 8:09 PM
dougm closed D20370: Eliminate code duplication around clip_start.
Thu, Jun 13, 8:09 PM
dougm updated the diff for D20625: clarify reserv_test_config.

Change bool literals to boolean_t literals in some places. Change u_ints to u_longs, and add a check for overflow. Truncate in computing hi.

Thu, Jun 13, 4:48 PM
dougm created D20628: Use vm_map_entry_t field prev less.
Thu, Jun 13, 3:04 AM

Wed, Jun 12

dougm updated the diff for D20625: clarify reserv_test_config.

Accept reviewer suggestions for changes to u_int for hi, lo and npages.

Wed, Jun 12, 8:44 PM
dougm added inline comments to D20370: Eliminate code duplication around clip_start.
Wed, Jun 12, 8:39 PM
dougm created D20625: clarify reserv_test_config.
Wed, Jun 12, 6:52 PM
dougm updated the diff for D20370: Eliminate code duplication around clip_start.

Restrict the scope of this patch just to reducing code duplication. On this amd64, it reduces the size of vm_map.o from 456608 to 455816, with no obvious impact on performance.

Wed, Jun 12, 6:23 PM
dougm added a comment to D20370: Eliminate code duplication around clip_start.

Ignore this patch while I study an unexpected negative performance impact.

Wed, Jun 12, 3:22 PM
dougm updated the diff for D20370: Eliminate code duplication around clip_start.

Restore a patch that got accidentally copied over last week.

Wed, Jun 12, 3:10 PM