Page MenuHomeFreeBSD

markj (Mark Johnston)
User

Projects

User Details

User Since
Mar 12 2014, 1:00 AM (247 w, 6 d)

Recent Activity

Today

markj committed rS341808: Remove an unused malloc(9) type..
Remove an unused malloc(9) type.
Tue, Dec 11, 2:16 AM
markj committed rS341807: Use inline tests for individual PTE bits in the RISC-V pmap..
Use inline tests for individual PTE bits in the RISC-V pmap.
Tue, Dec 11, 2:16 AM
markj closed D18461: Rename pmap_is_write() to pmap_l3_writable()..
Tue, Dec 11, 2:16 AM
markj added a comment to D18491: top(1): Add ability to sort by cpu core number.

I would suggest explaining the difference between -o cpu and -o core in the description of the -o option.

Tue, Dec 11, 2:11 AM
markj added a comment to D18342: Fix possible double free in bwn_{dma,pio}_tx_start()..

Sorry to nag, but did you get a chance to test this?

Tue, Dec 11, 1:30 AM

Yesterday

markj added inline comments to D18495: Fix sys.kern.coredump_phnum_test.coredump_phnum on i386.
Mon, Dec 10, 7:12 PM

Sun, Dec 9

markj updated the diff for D18492: Capsicumize savecore(8)..
  • Test.
Sun, Dec 9, 9:03 PM
markj updated the diff for D18492: Capsicumize savecore(8)..
  • Test.
Sun, Dec 9, 9:03 PM
markj updated the diff for D18492: Capsicumize savecore(8)..
  • Test.
Sun, Dec 9, 8:56 PM
markj updated the diff for D18492: Capsicumize savecore(8)..
  • Test
Sun, Dec 9, 8:56 PM
markj created D18492: Capsicumize savecore(8)..
Sun, Dec 9, 8:49 PM

Sat, Dec 8

markj updated the diff for D18458: Capsicumize savecore(8)..
  • Disable casper in the rescue build.
Sat, Dec 8, 6:22 PM

Fri, Dec 7

markj added inline comments to D18458: Capsicumize savecore(8)..
Fri, Dec 7, 10:07 PM
markj updated the diff for D18461: Rename pmap_is_write() to pmap_l3_writable()..
  • Get rid of the (inconsistently used) predicates and just use inline tests.
Fri, Dec 7, 10:00 PM
markj added inline comments to D18322: Expose threads-per-core and physical core count information.
Fri, Dec 7, 9:23 PM
markj added a comment to D18342: Fix possible double free in bwn_{dma,pio}_tx_start()..

This looks good to me; I'll try to test on real hardware over the weekend, just to be sure.

Fri, Dec 7, 8:21 PM
markj updated the diff for D18458: Capsicumize savecore(8)..
  • Add xfopenat().
Fri, Dec 7, 5:55 PM
markj added inline comments to D18458: Capsicumize savecore(8)..
Fri, Dec 7, 5:41 PM
markj updated the diff for D18458: Capsicumize savecore(8)..
  • Address feedback from cem.
Fri, Dec 7, 4:58 PM
markj updated subscribers of D18458: Capsicumize savecore(8)..
Fri, Dec 7, 4:58 PM
markj committed rS341694: Update the description of the address space layout on RISC-V..
Update the description of the address space layout on RISC-V.
Fri, Dec 7, 3:57 PM
markj closed D18463: Update the description of the address space layout on RISC-V..
Fri, Dec 7, 3:57 PM
markj committed rS341693: Rename sptbr to satp per v1.10 of the privileged architecture spec..
Rename sptbr to satp per v1.10 of the privileged architecture spec.
Fri, Dec 7, 3:57 PM
markj closed D18462: Rename sptbr to satp..
Fri, Dec 7, 3:57 PM
markj closed D18457: Let cap_syslog inherit stdio descriptors..
Fri, Dec 7, 3:54 PM
markj committed rS341692: Let the cap_syslog capability inherit stdio descriptors..
Let the cap_syslog capability inherit stdio descriptors.
Fri, Dec 7, 3:54 PM
markj committed rS341691: MFC r341346:.
MFC r341346:
Fri, Dec 7, 3:50 PM
markj added a comment to D17940: x86/acpica/srat.c: Add API for parsing proximity tables.

markj: Regarding the question on pxm prefix. I used the convention starting with the filename. I kind of settled for acpi_pxm since I could not think of a better filename and prefix (which covers SRAT for resource affinity and SLIT for locality).

An alternative would be to have acpi_proximity.c containing:
int acpi_parse_proximity_tables(int cpus, vm_paddr_t maxmem);
int acpi_set_mem_locality(void);
int acpi_set_cpu_locality(void);
void acpi_free_proximity_data(void);

Let me know if you have any suggestion on this.

Fri, Dec 7, 2:39 AM · arm64
markj updated the diff for D18464: Clean up initialization of phys_avail..
  • Further cleanup based on feedback.
Fri, Dec 7, 2:24 AM
markj added a comment to D18464: Clean up initialization of phys_avail..
In D18464#393352, @jhb wrote:

This is inspired by PR 231515?

Fri, Dec 7, 2:24 AM

Thu, Dec 6

markj added inline comments to D18461: Rename pmap_is_write() to pmap_l3_writable()..
Thu, Dec 6, 11:18 PM
markj added inline comments to D18461: Rename pmap_is_write() to pmap_l3_writable()..
Thu, Dec 6, 11:02 PM
markj updated the test plan for D18464: Clean up initialization of phys_avail..
Thu, Dec 6, 10:03 PM
markj updated the summary of D18463: Update the description of the address space layout on RISC-V..
Thu, Dec 6, 10:03 PM
markj updated the test plan for D18462: Rename sptbr to satp..
Thu, Dec 6, 10:03 PM
markj updated the test plan for D18461: Rename pmap_is_write() to pmap_l3_writable()..
Thu, Dec 6, 10:03 PM
markj created D18464: Clean up initialization of phys_avail..
Thu, Dec 6, 10:01 PM
markj created D18463: Update the description of the address space layout on RISC-V..
Thu, Dec 6, 10:01 PM
markj created D18462: Rename sptbr to satp..
Thu, Dec 6, 10:01 PM
markj created D18461: Rename pmap_is_write() to pmap_l3_writable()..
Thu, Dec 6, 10:00 PM
markj closed D18456: Add zdopen(), for use in capability mode..
Thu, Dec 6, 8:03 PM
markj committed rS341648: Add zdopen(3) to complement zopen(3)..
Add zdopen(3) to complement zopen(3).
Thu, Dec 6, 8:03 PM
markj added inline comments to D17773: UMA limits in zone & more.
Thu, Dec 6, 7:04 PM
markj updated the diff for D18456: Add zdopen(), for use in capability mode..

Update the zopen.3 man page.

Thu, Dec 6, 6:56 PM
markj updated the summary of D18456: Add zdopen(), for use in capability mode..
Thu, Dec 6, 5:41 PM
markj added reviewers for D18457: Let cap_syslog inherit stdio descriptors.: oshogbo, emaste, capsicum.
Thu, Dec 6, 5:41 PM
markj updated the summary of D18458: Capsicumize savecore(8)..
Thu, Dec 6, 5:41 PM
markj created D18458: Capsicumize savecore(8)..
Thu, Dec 6, 5:36 PM
markj created D18457: Let cap_syslog inherit stdio descriptors..
Thu, Dec 6, 5:36 PM
markj created D18456: Add zdopen(), for use in capability mode..
Thu, Dec 6, 5:36 PM
markj committed rS341638: Let kern.trap_enotcap be set as a tunable..
Let kern.trap_enotcap be set as a tunable.
Thu, Dec 6, 5:29 PM
markj accepted D18443: Fix bugs in plugable CC algorithm and siftr sysctls..
Thu, Dec 6, 4:55 PM
markj planned changes to D16667: Extend uma_reclaim() to permit different reclamation targets..
Thu, Dec 6, 4:30 PM
markj updated the diff for D16667: Extend uma_reclaim() to permit different reclamation targets..

Rebase.

Thu, Dec 6, 4:30 PM
markj accepted D18443: Fix bugs in plugable CC algorithm and siftr sysctls..
Thu, Dec 6, 4:29 PM
markj added inline comments to D18322: Expose threads-per-core and physical core count information.
Thu, Dec 6, 4:18 PM
markj added a comment to D18447: Temporarily skip flakey test cases.

Approved.

Thu, Dec 6, 3:44 PM
markj added inline comments to D18443: Fix bugs in plugable CC algorithm and siftr sysctls..
Thu, Dec 6, 4:27 AM
markj added inline comments to D18443: Fix bugs in plugable CC algorithm and siftr sysctls..
Thu, Dec 6, 12:52 AM

Wed, Dec 5

markj accepted D18062: gmirror: Evaluate mirror components against newest metadata copy.

@pho has a number of gmirror stress tests. Peter, do you mind giving this patch a try?

Wed, Dec 5, 7:53 PM
markj added a comment to D17773: UMA limits in zone & more.

Most of my inline comments are nits, but: I don't really understand how the wakeup() mechanism works with the per-CPU caches. Threads may remain asleep on "zonelimit" indefinitely if uz_items == uz_maxitems but active threads are able to avoid the slow path. Before, the ZFLAG_FULL check in the zfree fast path handled this. Is there anything preventing starvation?

Wed, Dec 5, 7:25 PM
markj committed rS341601: ixl(4) has netdump support via iflib..
ixl(4) has netdump support via iflib.
Wed, Dec 5, 6:22 PM
markj accepted D18062: gmirror: Evaluate mirror components against newest metadata copy.
Wed, Dec 5, 5:26 PM
markj added inline comments to D18322: Expose threads-per-core and physical core count information.
Wed, Dec 5, 5:19 PM
markj closed D17803: Clamp porthash_max to IPPORT_MAX..
Wed, Dec 5, 5:09 PM
markj committed rS341595: Clamp the INPCB port hash tables to IPPORT_MAX + 1 chains..
Clamp the INPCB port hash tables to IPPORT_MAX + 1 chains.
Wed, Dec 5, 5:09 PM
markj committed rS341594: Netdump support was added to mlx4en(4) in r341548..
Netdump support was added to mlx4en(4) in r341548.
Wed, Dec 5, 4:58 PM

Tue, Dec 4

markj committed rS341491: MFC r341442, r341443:.
MFC r341442, r341443:
Tue, Dec 4, 7:07 PM
markj committed rS341490: MFC r341442, r341443:.
MFC r341442, r341443:
Tue, Dec 4, 7:06 PM
markj accepted D17940: x86/acpica/srat.c: Add API for parsing proximity tables.
Tue, Dec 4, 5:05 PM · arm64
markj added inline comments to D18359: getfhat, fhlink, fhlinkat, fhreadlink: new file handle system calls.
Tue, Dec 4, 5:02 PM · manpages
markj updated the diff for D17803: Clamp porthash_max to IPPORT_MAX..
  • Fix an off-by-one.
  • Retire INP_PCBLBGROUP_PORTHASH. It's identical to INP_PCBPORTHASH, and because we now size the hash tables accurately, it would be nice to consistently use the same port hash everywhere.
Tue, Dec 4, 4:34 PM
markj updated the summary of D17226: Implement M_NEXTFIT..
Tue, Dec 4, 3:25 PM
markj added inline comments to D17226: Implement M_NEXTFIT..
Tue, Dec 4, 12:54 AM
markj updated the diff for D17226: Implement M_NEXTFIT..
  • Attempt to handle all of the bugs that Alan pointed out.
Tue, Dec 4, 12:54 AM
markj added inline comments to D17773: UMA limits in zone & more.
Tue, Dec 4, 12:33 AM

Mon, Dec 3

markj added inline comments to D17773: UMA limits in zone & more.
Mon, Dec 3, 11:46 PM
markj added inline comments to D5603: ASLR.
Mon, Dec 3, 10:02 PM
markj accepted D5603: ASLR.

I suspect that the anon clustering will interact suboptimally with the jemalloc behaviour discussed in D16501 and elsewhere. In particular, jemalloc will unmap small regions of the address space, leaving holes. With clustering, those holes won't be reused since we no longer perform a first-fit search. IMO it would be worth reconsidering how anon_loc works; rather than advancing it after each successful clustering, maybe it should be constant after the initialization to a non-zero value, so that we attempt to fill holes with new mappings before extending the clustered region further. I do not think this needs to be done prior to commit though.

Mon, Dec 3, 9:44 PM
markj committed rS341443: Add a missing definition for the !COMPAT_FREEBSD32 case..
Add a missing definition for the !COMPAT_FREEBSD32 case.
Mon, Dec 3, 9:08 PM
markj committed rS341442: Plug memory disclosures via ptrace(2)..
Plug memory disclosures via ptrace(2).
Mon, Dec 3, 8:54 PM
markj closed D18421: Fix memory disclosures from ptrace(PT_GET*REGS)..
Mon, Dec 3, 8:54 PM
markj added a comment to D18362: Temporarily skip flip-flop test cases.

Approved

Mon, Dec 3, 8:47 PM
markj added inline comments to D5603: ASLR.
Mon, Dec 3, 8:40 PM
markj updated the diff for D17854: Add cap_sysctl() and cap_sysctlnametomib()..

Address some of the feedback. I left replies for the rest.

Mon, Dec 3, 7:43 PM
markj updated the diff for D17572: Run most of rtsold(8) and rtsol(8) in capability mode..

Address remaining issues:

  • Generalize the probe service to all unconnected sendmsg() calls. I had been creating one pre-connected socket per interface, but that only works if the interface is up when rtsold starts. If it comes up after we enter capability mode, we're stuck.
  • We can't call getifaddrs() in capability mode. (The SIOCGIFINFO ioctl does effectively the same thing, though, and is permitted in capability mode.) For now, just add yet another casper service to get the address flags of the link-local address on a specified interface.
Mon, Dec 3, 7:40 PM
markj updated the diff for D18421: Fix memory disclosures from ptrace(PT_GET*REGS)..

Explicitly bzero 32bit compat register structures in sys_ptrace().

Mon, Dec 3, 5:45 PM
markj updated the diff for D18421: Fix memory disclosures from ptrace(PT_GET*REGS)..
  • Zero out reg structs in procfs too.
Mon, Dec 3, 5:42 PM
markj added inline comments to D18421: Fix memory disclosures from ptrace(PT_GET*REGS)..
Mon, Dec 3, 5:27 PM
markj added reviewers for D18421: Fix memory disclosures from ptrace(PT_GET*REGS).: jhb, kib.
Mon, Dec 3, 4:35 PM
markj created D18421: Fix memory disclosures from ptrace(PT_GET*REGS)..
Mon, Dec 3, 4:29 PM
markj committed rS341426: MFC r341001:.
MFC r341001:
Mon, Dec 3, 3:16 PM
markj added a comment to D17940: x86/acpica/srat.c: Add API for parsing proximity tables.

This mostly LGTM, a couple of nits inline.

Mon, Dec 3, 3:13 PM · arm64
markj accepted D17941: acpica : move SRAT/SLIT parsing to sys/dev/acpica.
Mon, Dec 3, 2:51 PM
markj added inline comments to D17773: UMA limits in zone & more.
Mon, Dec 3, 2:47 PM

Sun, Dec 2

markj committed rS341401: MFC r341247:.
MFC r341247:
Sun, Dec 2, 6:09 PM

Sat, Dec 1

markj committed rS341376: MFC r340861, r340900:.
MFC r340861, r340900:
Sat, Dec 1, 5:48 PM

Fri, Nov 30

markj added a reviewer for D18395: Add uk.macbook.kbd keymap (vt): emaste.
Fri, Nov 30, 6:22 PM · Contributor Reviews (base)
markj committed rS341346: Add missing display messages when toggling modes..
Add missing display messages when toggling modes.
Fri, Nov 30, 6:04 PM
markj closed D18331: top(1): Add missing display messages when toggling modes.
Fri, Nov 30, 6:04 PM