Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_567589295 @arozansk @ddutile @llong1 It would be great if you could have another look at v3 -- thanks! ___ kernel mailing list --

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: Rafael Aquini on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_558913497 Thanks David, that's fair enough. Works with me! ___ kernel mailing list -- kernel@lists.fedoraproject.org To unsubscribe send an email to

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_558576985 I consider it as helpful on ppc64 and s390x as well to troubleshoot, just to know that CMA allocations happened. After all, it's a single pr_info() we'll spot in dmesg. However,

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: Rafael Aquini on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_558452775 David, minor question: don't we want to follow the same !IS_ENABLED(CONFIG_S390) && !IS_ENABLED(CONFIG_PPC64) check here, as well? ___

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_556204600 Yes, that's what the current patches do (feel free to review and eventually ack ;) ) Thanks! ___ kernel mailing list --

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: Rafael Aquini on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_556194632 I don't have a strong opinion on this one, but I think should perhaps go with (albeit weird) the conditional tainting for x86_64 and Aarch64. Reason CMA was disabled in RHEL8

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_554101592 Right, if the user manually specifies "cma=2G" and the system is not able to cope with that, this is a clear configuration error and once supported, we would want to document that

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: Dave Young on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_554095592 Hi David, Ok, I had a discussion in kdump team, for X86 since we have crashkernel=,high and we can reserve more in high memory region so the risk is low. The only case is what you

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_552329069 Sure, but keep in mind that it will be tech preview and is similarly enabled already in Fedora. However, you remind me of an interesting case: kdump and CMA reservations We would

Re: [OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: Dave Young on gitlab.com https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023#note_552056585 There might be some side effects to kdump crashkernel reservation. Kdump team need to do some test quickly first. ___ kernel mailing list --

[OS-BUILD PATCHv2 0/2] enable CONFIG_CMA on x86_64 in ARK

From: David Hildenbrand on gitlab.com Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1023 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1945002 Enable CONFIG_CMA on x86_64 in ARK as we already have in fedora; CMA is Tech Preview in RHEL 9.