It’s a bug alright — in the kernel. ioctl’s are done on files that have already been opened, there’s no way in hell that ENOENT would ever be valid.
Shut up, Mauro. And you’ve shown yourself to not be competent in this issue, so I’ll apply it directly and immediately myself.
WE DO NOT BREAK USERSPACE!
Seriously. And fix your approach to kernel programming.
Linus — To unsubscribe from this list: send the line «unsubscribe linux-kernel» in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
«>Re: [Regression w/ patch] Media commit causes user.
Dan Li
[PATCH] [PATCH V2]ARM64: SCS: Add gcc plugin to su.
Ansuel Smith
We have a bit to set the mode. I tried to implement this as we currently qca8k is hardcoded to expect the cpu port 0 for everything and doesn’t actually found a valid cpu port (aka it doesn’t expect a configuration with cpu6) I think the driver was writtent with the concept of mac exchange from the start. So I think I will just drop the mac exchange and fix the code to make it dynamic. exchange and fix the code to make it dynamic. «>Re: [net-next PATCH v2 08/15] dt-bindings: net: ds.
while ago), and that’s the one that blew up. It has nothing to do with the config (the same config compiles fine with gcc 10.x). And if I didn’t have a test that compiled with 8.1.0, I would never had know this was an issue. would never had know this was an issue. workarounds that actively lie to the compiler. I felt uncomfortable with the change, and that’s why I wanted to get your opinion before having you first see it in a pull request. «>Re: [RFC][PATCH] tracing: Define «fake» struct tra.
«Kirill A. Shutemov»
Compaction at the moment built with assumption that compound pages are PMD-mappable or larger and it doesn’t make sense to move them: PMD-mappable or larger and it doesn’t make sense to move them: /* * Regardless of being on LRU, compound pages such as THP and * hugetlbfs are not to be compacted unless we are attempting * an allocation much larger than the huge page size (eg CMA). «>Re: Compaction & folios
LKML.ORG?
In case you haven’t read the titlebar of your webbrowser’s window: this site is the (unofficial) Linux Kernel Mailing List archive. This mailing list is a rather high-volume list, where (technical) discussions on the design of, and bugs in the Linux kernel take place. If that scares you, please read the FAQ.
This is pretty confusing. I don’t see how the patch could cause this to happen.
yeah. Please check the return value and print an error
diff —git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c index 1800053b4644..3c4538e26731 100644 — a/drivers/clocksource/exynos_mct.c +++ b/drivers/clocksource/exynos_mct.c @@ -450,6 +450,7 @@ static int exynos4_local_timer_setup(struct clock_event_device *evt)
Jan Kara
Also since you call ->write_inode directly upper->i_state won’t be updated to reflect that inode has been written out (I_DIRTY flags get cleared in __writeback_single_inode()). So it seems to me overlayfs will keep writing out upper inode until flush worker on upper filesystem also writes the inode and clears the dirty flags? So you rather need to call something like write_inode_now() that will handle the flag clearing and do writeback list handling for you? handling for you? Honza Honza «>Re: [RFC PATCH v5 06/10] ovl: implement overlayfs’.
Jason Wang
0 by default. 0 by default. I think 0 is not a correct value. If I understand the spec correctly, it should be 1. should be 1. Thanks Thanks Thanks VDPA_ATTR_MAX, VDPA_ATTR_MAX, «>Re: [PATCH v4 6/7] vdpa: add new attribute VDPA_AT.
«Rafael J. Wysocki»
Hi Linus, Please pull from the tag Please pull from the tag git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \ acpi-5.15-rc5 acpi-5.15-rc5 with top-most commit 3fb937f441c64af1eec60bfd3732f64001fcc534 with top-most commit 3fb937f441c64af1eec60bfd3732f64001fcc534 PCI: ACPI: Check parent pointer in acpi_pci_find_companion() PCI: ACPI: Check parent pointer in acpi_pci_find_companion() on top of commit 9e1ff307c779ce1f0f810c7ecce3d95bbae40896 on top of commit 9e1ff307c779ce1f0f810c7ecce3d95bbae40896 Linux 5.15-rc4 Linux 5.15-rc4 to receive an ACPI fix for 5.15-rc5. to receive an ACPI fix for 5.15-rc5. This fixes a recent ACPI-related regression in the PCI subsystem that introduced a NULL pointer dereference possible to trigger from user space via sysfs on some systems. from user space via sysfs on some systems. Thanks! Thanks! Thanks! ————— Rafael J. «>[GIT PULL] ACPI fix for v5.15-rc5
q-mailbox.c ructions + * instruction time out 2^22*2*6.41ns =3D 53ms For different clients, the timeout value would be different, and each client could use timer to detect timeout, so it’s not necessary to enable timeout in cmdq driver. enable timeout in cmdq driver. Regards, Chun-Kuang. Chun-Kuang. , void *data) _pa, _INSTN_TIMEOUT_CYCLES); ); E); ASK); 2.25.1 2.25.1 «>Re: [PATCH v2, 1/1] mailbox: cmdq: add instruction.
Andy Shevchenko
[PATCH v3 1/1] gpio: mockup: Convert to use softwa.
«Zhang, Yanmin»
[PATCH V5 3/3] perf & kvm: Enhance perf to collect.
allocations in this same distribution. Which means that any time an order n allocation fails, it’s likely that we’ll still have order n-1 pages free — and of those free order n-1 pages, one will likely have a buddy that’s moveable and hasn’t been fragmented — meaning the common case is that compaction will have to move _one_ (higher order) page — we’ll almost never be having to move a bunch of 4k pages. «>Compaction & folios
Borislav Petkov
Hi Linus, the last two weeks gathered in several x86/urgent fixes, please pull them in. them in. Thx. Thx. — The following changes since commit 5816b3e6577eaa676ceb00a848f0fd65fe2adc29: Linux 5.15-rc3 (2021-09-26 14:08:19 -0700) are available in the Git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git tags/x86_urgent_for_v5.15_rc5 for you to fetch changes up to d298b03506d3e161f7492c440babb0bfae35e650: x86/fpu: Restore the masking out of reserved MXCSR bits (2021-10-08 11:12:17 +0200) —————————————————————- — A FPU fix to properly handle invalid MXCSR values: 32-bit masks them out due to histerical reasons and 64-bit kernels reject them — A fix to clear X86_FEATURE_SMAP when support for is not config-enabled — Three fixes correcting misspelled Kconfig symbols used in code — Two resctrl object cleanup fixes — Yet another attempt at fixing the neverending saga of botched x86 timers, «>[GIT PULL] x86/urgent for v5.15-rc5
corkflag af_unix: fix races in sk_peer_pid and sk_peer_cred accesses Eric Whitney (2): ext4: remove extent cache entries when truncating inline data ext4: enforce buffer head state assertion in ext4_da_map_blocks Evgeny Novik»>Linux 5.15-rc4
Tao Zhou
wake_avg_idle; avg_cost = this_sd->avg_scan_cost + 1; — span_avg = sd->span_weight * avg_idle; + span_avg = this_sd->span_weight * avg_idle; 4*avg_cost) nr = div_u64(span_avg, avg_cost); «>[PATCH] sched/fair: Use this_sd->weight to calcula.
Matthew Wilcox
> > > > Secondly, there is no need to check if folio is pointer or value with ‘xa_is_value(folio)’ > because folio is alwayse pointer. > > > > xa_is_value(folio)) > + if (!folio) > > FGP_HEAD)) > It does pointer arithmetic on folio. Specifically, it adds zero to folio, because page is the first element of folio. FGP_ENTRY) return folio; «>Re: [PATCH linux-next] mm/folio-compat: fix potent.
Peter Zijlstra
6. completely destroys and ignores any machine topology information. 6. completely destroys and ignores any machine topology information. 6. completely destroys and ignores any machine topology information. «>Re: [RFC 0/5] kernel: Introduce CPU Namespace
Hyeonggon Yoo
L1 Data Cache: 32KB, 8-way associative, 64 byte line size L1 Instruction Cache: 32KB, 8-way associative, 64 byte line size L2 Unified Cache: 1MB, 16-way associative, 64 byte line size L3 Unified Cache: 33MB, 11-way associative, 64 byte line size L3 Unified Cache: 33MB, 11-way associative, 64 byte line size L3 Unified Cache: 33MB, 11-way associative, 64 byte line size I run hackbench with per-node coloring, per-cpu coloring, and without coloring. coloring. hackbench -g 100 -l 200000 without coloring: 2196.787 with per-node coloring: 2193.607 with per-cpu coloring: 2198.076 it seems there is almost no difference. «>Almost no difference
Matthew Wilcox
single TLB entry to cache an entire 2MB page. But there are the software ones, which I believe Google have measured (perhaps it was the ChromeOS team?) Allocating order-2/3/4 pages reduces the length of the LRU list by a factor of 4/8/16. «>Re: Compaction & folios
Linus Torvalds
Mauro, SHUT THE FUCK UP!
It’s a bug alright — in the kernel. ioctl’s are done on files that have already been opened, there’s no way in hell that ENOENT would ever be valid.
Shut up, Mauro. And you’ve shown yourself to not be competent in this issue, so I’ll apply it directly and immediately myself.
WE DO NOT BREAK USERSPACE!
Seriously. And fix your approach to kernel programming.
Linus — To unsubscribe from this list: send the line «unsubscribe linux-kernel» in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
«>Re: [Regression w/ patch] Media commit causes user.
Dan Li
[PATCH] [PATCH V2]ARM64: SCS: Add gcc plugin to su.
Ansuel Smith
We have a bit to set the mode. I tried to implement this as we currently qca8k is hardcoded to expect the cpu port 0 for everything and doesn’t actually found a valid cpu port (aka it doesn’t expect a configuration with cpu6) I think the driver was writtent with the concept of mac exchange from the start. So I think I will just drop the mac exchange and fix the code to make it dynamic. exchange and fix the code to make it dynamic. «>Re: [net-next PATCH v2 08/15] dt-bindings: net: ds.
while ago), and that’s the one that blew up. It has nothing to do with the config (the same config compiles fine with gcc 10.x). And if I didn’t have a test that compiled with 8.1.0, I would never had know this was an issue. would never had know this was an issue. workarounds that actively lie to the compiler. I felt uncomfortable with the change, and that’s why I wanted to get your opinion before having you first see it in a pull request. «>Re: [RFC][PATCH] tracing: Define «fake» struct tra.
«Kirill A. Shutemov»
Compaction at the moment built with assumption that compound pages are PMD-mappable or larger and it doesn’t make sense to move them: PMD-mappable or larger and it doesn’t make sense to move them: /* * Regardless of being on LRU, compound pages such as THP and * hugetlbfs are not to be compacted unless we are attempting * an allocation much larger than the huge page size (eg CMA). «>Re: Compaction & folios
own edge configuration. But that then leads into the MAC0/MAC6 swap mess. I need to think about that some more, how do we cleanly describe that in DT. Andrew Andrew Andrew «>Re: [net-next PATCH v2 08/15] dt-bindings: net: ds.
Saurav Girepunje
I can do that in either one of the way. Either I should check for ret before free or I should use goto out(may be I will rename err_out goto label to «out» only) just after the memcpy statement so that function will not free for no error case and return ret value only. ret value only. Please let me know you input on above. Please let me know you input on above. Regards, Saurav Girepunje Saurav Girepunje Saurav Girepunje «>Re: [PATCH] staging: rtl8723bs: os_dep: simplify t.
Ira Weiny
base — map->block_offset); But I think map->base should be ‘base’ and map->block_offset should be handled in cxl_probe_regs() rather than subtract it here.. subtract it here.. Either way this is cleaner than what it was. Either way this is cleaner than what it was. Reviewed-by: Ira Weiny Reviewed-by: Ira Weiny return ret; return ret; «>Re: [PATCH v3 06/10] cxl/pci: Add @base to cxl_reg.
Paul Cercueil
oob_poi — * @page: page number to read — * — * Hardware ECC for large page chips, require OOB to be read first. NAND_ECC_GENERIC_ERASED_CHECK)) < - /* check for empty pages with bitflips */ - stat = nand_check_erased_ecc_chunk(p, eccsize, - &ecc_code[i], - eccbytes, NULL, 0, - chip->ecc.strength); «>[PATCH 2/3] mtd: rawnand: Export nand_read_page_hw.
Hottest messages — half-life = 1 month
Linus Torvalds
30 years since the Linux 0.01 release
Linus Torvalds
the new threat. And of the vaccines, the mRNA ones are the most modern, and the most targeted — exactly because they do *not* need to have any of the other genetic material that you traditionally have in a vaccine (ie no need for basically the whole — if weakened — bacterial or virus genetic material). Yes, all those vaccinated people around you will protect you too, but if there is another wave, possibly due to a more transmissible version — you and your family will be at _much_ higher risk than those vaccinated people because of your ignorance and mis-information. mis-information. Get vaccinated. «>Re: Maintainers / Kernel Summit 2021 planning kick.
Qu Wenruo
Hi Leizhen, and guys in the mail list, Recently I find one patch removing a debug OOM error message from btrfs=20 selftest. selftest. It’s nothing special, some small cleanup work from some kernel newbie. After some quick search, more and more oom error message «cleanup»=20 patches just show up, even some misspell fixes. patches just show up, even some misspell fixes. patches just show up, even some misspell fixes. It’s OK for first-time/student developers to submit such patches, and I=20 really hope such patches would make them become a long term contributor. «>Please don’t waste maintainers’ time on your KPI g.
Linus Torvalds
integers or floating point». So if the Rust compiler causes hidden allocations that cannot be caught and returned as errors, then I seriously think that this whole approach needs to be entirely NAK’ed, and the Rust infrastructure — whether at the compiler level or in the kernel wrappers — needs more work. work. So if the panic was just some placeholder for things that _can_ be caught, then I think that catching code absolutely needs to be written, and not left as a to-do. written, and not left as a to-do. And if the panic situation is some fundamental «this is what the Rust compiler does for internal allocation failures», then I think it needs more than just kernel wrapper work — it needs the Rust compiler to be *fixed*. *fixed*. Because kernel code is different from random user-space system tools. «>Re: [PATCH 00/13] [RFC] Rust support
[email protected] .
[PATCH 00/13] [RFC] Rust support
Linus Torvalds
Mauro, SHUT THE FUCK UP!
It’s a bug alright — in the kernel. ioctl’s are done on files that have already been opened, there’s no way in hell that ENOENT would ever be valid.
Shut up, Mauro. And you’ve shown yourself to not be competent in this issue, so I’ll apply it directly and immediately myself.
WE DO NOT BREAK USERSPACE!
Seriously. And fix your approach to kernel programming.
