drop support for USE=32bit_au |
||||||
Issue descriptionthese devices transitioned to 64-bit in the following releases. first date is for the CrOS release (not the device). the EOL date is estimated, not final. lumpy R19 May 2012 (EOL: May 2017) stumpy R19 May 2012 (EOL: Jul 2017) parrot R32 Jan 2014 (EOL: Oct 2017) butterfly R33 Feb 2014 (EOL: Feb 2018) stout R36 Jul 2014 (EOL: Dec 2017) if we have omaha rules in place to force a stepping version for these boards at or after that release, then we no longer need the 32-bit upgrade path in the installer, and we can drop a decent bit of code that is causing us toolchain pain atm. seems like having a stepping stone in place for all devices at R37 shouldn't be a big deal ... that's almost 3 years ago at this point, so the number of devices that have to upgrade from <R37 should be significantly low. going to assign to Bernie first to help locate stepping stone versions. feel free to bounce between TPMs/whoever.
,
Jul 26 2017
sounds like it, thanks! i'll post CLs to start dropping stuff.
,
Jul 27 2017
The following revision refers to this bug: https://chrome-internal.googlesource.com/chromeos/overlays/overlay-stout-private/+/61a21b8a985b2fe1733fa4335967ec03fbcb962a commit 61a21b8a985b2fe1733fa4335967ec03fbcb962a Author: Mike Frysinger <vapier@chromium.org> Date: Thu Jul 27 05:41:53 2017
,
Jul 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/board-overlays/+/ebf9a38acb1ab7af7313c7f6d80530dd0c5a1b8a commit ebf9a38acb1ab7af7313c7f6d80530dd0c5a1b8a Author: Mike Frysinger <vapier@chromium.org> Date: Thu Jul 27 05:41:57 2017 butterfly: drop USE=32bit_au support Butterfly changed to pure 64-bit starting in R33. We have a stepping stone rule for butterfly so that it upgrades to R53 before moving to newer versions, and that includes 32bit_au support. So we don't need this logic in newer versions anymore. BUG= chromium:749166 TEST=precq passes Change-Id: I545851175eb55e63b90e454912b489f523fdcda4 Reviewed-on: https://chromium-review.googlesource.com/587788 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/ebf9a38acb1ab7af7313c7f6d80530dd0c5a1b8a/overlay-butterfly/profiles/base/make.defaults
,
Jul 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/board-overlays/+/0579d93b9c0b73f365e81a0904b4e7d626adfdce commit 0579d93b9c0b73f365e81a0904b4e7d626adfdce Author: Mike Frysinger <vapier@chromium.org> Date: Thu Jul 27 05:41:57 2017 parrot: drop USE=32bit_au support Parrot changed to pure 64-bit starting in R32. We have a stepping stone rule for parrot so that it upgrades to R53 before moving to newer versions, and that includes 32bit_au support. So we don't need this logic in newer versions anymore. BUG= chromium:749166 TEST=precq passes Change-Id: I0020ee9b3f02b2df837429a4ddfb0234e6e55df8 Reviewed-on: https://chromium-review.googlesource.com/587789 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/0579d93b9c0b73f365e81a0904b4e7d626adfdce/overlay-parrot/profiles/base/make.defaults
,
Jul 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/board-overlays/+/eefbe4759bf7f6389b05c5c87fc849aebfbb629e commit eefbe4759bf7f6389b05c5c87fc849aebfbb629e Author: Mike Frysinger <vapier@chromium.org> Date: Thu Jul 27 05:41:54 2017 lumpy: drop USE=32bit_au support Lumpy changed to pure 64-bit starting in R19. We have a stepping stone rule for lumpy so that it upgrades to R23 before moving to newer versions, and that includes 32bit_au support. So we don't need this logic in newer versions anymore. BUG= chromium:749166 TEST=precq passes Change-Id: Id9f2266f0fa627ab87b5472feddd3e4e6c1f06fe Reviewed-on: https://chromium-review.googlesource.com/587887 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/eefbe4759bf7f6389b05c5c87fc849aebfbb629e/overlay-lumpy/profiles/base/make.defaults
,
Jul 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/board-overlays/+/f4e76a3f9e15ff6f81cd78d0f6c8443f12c237fe commit f4e76a3f9e15ff6f81cd78d0f6c8443f12c237fe Author: Mike Frysinger <vapier@chromium.org> Date: Thu Jul 27 05:41:53 2017 stumpy: drop USE=32bit_au support Stumpy changed to pure 64-bit starting in R19. We have a stepping stone rule for stumpy so that it upgrades to R23 before moving to newer versions, and that includes 32bit_au support. So we don't need this logic in newer versions anymore. BUG= chromium:749166 TEST=precq passes Change-Id: I5be666d4f09b1af8e9e0636b101937b7ffd25767 Reviewed-on: https://chromium-review.googlesource.com/587888 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/f4e76a3f9e15ff6f81cd78d0f6c8443f12c237fe/overlay-stumpy/profiles/base/make.defaults
,
Jul 28 2017
it's now off for all boards. i'll hold on landing the ebuild cleanups for a bit to make sure we don't need to revert for some random reason. ideally we'd even wait until after the next branch (R62), but i'm guessing the toolchain guys would rather move quicker than that ;).
,
Aug 1 2017
yes, please. Thanks again.
,
Aug 4 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/a9d1856f04ba2b79885fe93698e7ccaca6544ef3 commit a9d1856f04ba2b79885fe93698e7ccaca6544ef3 Author: Mike Frysinger <vapier@chromium.org> Date: Fri Aug 04 00:11:01 2017 chromeos-installer: drop USE=32bit_au support Now that no boards use this flag, we can drop support for it. BUG= chromium:749166 TEST=precq passes Change-Id: Ib818e35ec8abbfff8d8dae050ff574b133ccab90 Reviewed-on: https://chromium-review.googlesource.com/587949 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/a9d1856f04ba2b79885fe93698e7ccaca6544ef3/chromeos-base/chromeos-installer/chromeos-installer-9999.ebuild
,
Aug 8 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/6c2c831e6f090a24001afeed7e4673444110206f commit 6c2c831e6f090a24001afeed7e4673444110206f Author: Mike Frysinger <vapier@chromium.org> Date: Tue Aug 08 17:09:12 2017 vboot_reference: drop USE=32bit_au support Now that no boards use this flag, we can drop support for it. BUG= chromium:749166 TEST=precq passes Change-Id: I974e13d533c3f979daef956767ae3d2c40dac094 Reviewed-on: https://chromium-review.googlesource.com/587948 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [modify] https://crrev.com/6c2c831e6f090a24001afeed7e4673444110206f/chromeos-base/vboot_reference/vboot_reference-9999.ebuild
,
Aug 9 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/89d389c35f71c3b33607d70f5f25929f81deaaa4 commit 89d389c35f71c3b33607d70f5f25929f81deaaa4 Author: Mike Frysinger <vapier@chromium.org> Date: Wed Aug 09 21:05:07 2017 gmock32/gtest32: drop unused packages Now that no packages depend on these, we can drop them. BUG= chromium:749166 TEST=precq passes Change-Id: Ib055a0a68a6b7299c6b6fa2003fe842df4cd3185 Reviewed-on: https://chromium-review.googlesource.com/587950 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Luis Lozano <llozano@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gtest32/gtest32-1.8.0.ebuild [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gtest32/Manifest [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gmock32/gmock32-1.8.0.ebuild [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/metadata/md5-cache/dev-cpp/gmock32-1.7.0-r1 [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gtest32/files/gtest-1.8.0-makefile-am.patch [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gmock32/files/gmock-1.8.0-makefile-am.patch [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gmock32/files/gtest-config [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gmock32/Manifest [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/dev-cpp/gtest32/files/configure-fix-pthread-linking.patch [delete] https://crrev.com/52094a8860051c71fa1b207ac1443252e9a6b415/metadata/md5-cache/dev-cpp/gtest32-1.7.0
,
Aug 9 2017
i have one more CL to drop the cros-au eclass and we're done with scrubbing this functionality however, it doesn't get rid of x86/32-bit support entirely -- the cros-i686 eclass is still around because breakpad has limitations (and crash-reporter is beholden to that). i've opened a new bug to track that: https://bugs.chromium.org/p/google-breakpad/issues/detail?id=740 i don't expect to work on it any time soon though. note: i don't expect upgrading ARC++ to a native 64-bit env will help. it'll still have 32-bit runtime support that apps can leverage which means our crash framework will need to support it.
,
Aug 14 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/3c431c0f3650d6caaa147c82d46bb3754496990f commit 3c431c0f3650d6caaa147c82d46bb3754496990f Author: Mike Frysinger <vapier@chromium.org> Date: Mon Aug 14 16:18:43 2017 cros-au: punt eclass The only user of this code is the cros-i686 eclass, so move the funcs into that so we can drop cros-au entirely. BUG= chromium:749166 TEST=precq passes Change-Id: Ibbb6f5006ade181b109ac260aa1c6295fcbd2825 Reviewed-on: https://chromium-review.googlesource.com/609035 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Luis Lozano <llozano@chromium.org> [delete] https://crrev.com/cb5310e6b5dba62f3a06ee4add779352fd8944ac/eclass/cros-au.eclass [modify] https://crrev.com/3c431c0f3650d6caaa147c82d46bb3754496990f/eclass/cros-i686.eclass
,
Sep 26 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/platform/firmware/+/0c73d3778abcfc8fff7bbea1083bf682ebbd8f1b commit 0c73d3778abcfc8fff7bbea1083bf682ebbd8f1b Author: Mike Frysinger <vapier@chromium.org> Date: Tue Sep 26 10:29:22 2017 drop support for 32-bit fallback binaries We don't have any 32->64 bit transitions anymore, so drop support for it. BUG= chromium:749166 TEST=precq passes Change-Id: I831779089a52cc3ec9ef4b7e59fd9f0bd213c2e3 Reviewed-on: https://chromium-review.googlesource.com/668659 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Hung-Te Lin <hungte@chromium.org> Reviewed-by: Don Garrett <dgarrett@chromium.org> [modify] https://crrev.com/0c73d3778abcfc8fff7bbea1083bf682ebbd8f1b/pack_dist/crosutil.sh
,
Dec 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/overlays/chromiumos-overlay/+/7610755a0e83b99999e89148faa217c050723359 commit 7610755a0e83b99999e89148faa217c050723359 Author: Mike Frysinger <vapier@chromium.org> Date: Wed Dec 06 19:33:09 2017 chromeos-installer: drop USE=32bit_au support Now that no boards use this flag, we can drop support for it. BUG= chromium:749166 TEST=precq passes Change-Id: Ib818e35ec8abbfff8d8dae050ff574b133ccab90 Reviewed-on: https://chromium-review.googlesource.com/587949 Commit-Ready: Mike Frysinger <vapier@chromium.org> Tested-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Bernie Thompson <bhthompson@chromium.org> (cherry picked from commit a9d1856f04ba2b79885fe93698e7ccaca6544ef3) Signed-off-by: Gwendal Grignou <gwendal@chromium.org> Reviewed-on: https://chromium-review.googlesource.com/809524 [modify] https://crrev.com/7610755a0e83b99999e89148faa217c050723359/chromeos-base/chromeos-installer/chromeos-installer-9999.ebuild
,
Jan 22 2018
,
Jun 21 2018
|
||||||
►
Sign in to add a comment |
||||||
Comment 1 by bhthompson@google.com
, Jul 26 2017For Lumpy I think we are already doing this: # Stepping stone rule to keep Full update from pre-R23 lumpy lumpy 1412.245.0 -> 2913.331.0 Rule { Update { UpdatedVersion: "2913.331.0" UrlBase: "https://dl.google.com/chromeos/lumpy/2913.331.0/stable-channel" PackageItem: { Name: "chromeos_2913.331.0_lumpy_stable-channel_full_mp-v2.bin-021e0b0176602796ab39db90b60564b9.signed" Size: "381139237" Hash: "hbbZjfK/vBanETg5rXLfuTVrxeI=" } Pair { Tag: "ChromeVersion" Value: "23.0.1271.111" } Pair { Tag: "ChromeOSVersion" Value: "{updated_version}" } Pair { Tag: "MaxDaysToScatter" Value: "14" } Pair { Tag: "MetadataSize" Value: "48735" } Pair { Tag: "MetadataSignatureRsa" Value: "" } Pair { Tag: "sha256" Value: "9V5HRq0qaEohUPmISF29sXCFvFg6zED1hDWYv4qc2ZM=" } Pair { Tag: "IsDelta" Value: "true" } Pair { Tag: "IsDeltaPayload" Value: "false" } Pair { Tag: "deadline" Value: "now" } } Test { Version: "[1412.245.0-{updated_version})" } Test { Attr: "hardware_class" Match: "^LUMPY .*" } Test { Attr: "track" Match: "stable-channel" } Fraction: 1.0 } Stumpy has a few, but should also be beyond this point: # Stepping stone rule to keep Full update from older that R23 versions -> 2913.331.0 Rule { Update { UpdatedVersion: "2913.331.0" UrlBase: "https://dl.google.com/chromeos/stumpy/2913.331.0/stable-channel" PackageItem: { Name: "chromeos_2913.331.0_stumpy_stable-channel_full_mp-v2.bin-f1b6e1e31ed7091e353969381e0f773d.signed" Size: "374215338" Hash: "YxVAhY+zsw7SAyRIE3WjihacO1Q=" } Pair { Tag: "ChromeVersion" Value: "23.0.1271.111" } Pair { Tag: "ChromeOSVersion" Value: "{updated_version}" } Pair { Tag: "MaxDaysToScatter" Value: "14" } Pair { Tag: "MetadataSize" Value: "48687" } Pair { Tag: "MetadataSignatureRsa" Value: "" } Pair { Tag: "sha256" Value: "1adVKdPKoanJXr94dAnKRK11lqLnI13+3ryaS/TohAQ=" } Pair { Tag: "IsDelta" Value: "true" } Pair { Tag: "IsDeltaPayload" Value: "false" } Pair { Tag: "deadline" Value: "now" } } Test { Version: "[1412.205.2-{updated_version})" } Test { Attr: "hardware_class" Match: "^STUMPY .*" } Test { Attr: "track" Match: "stable-channel" } Fraction: 1.0 } Parrot has a 53 one that I think will achieve this: # Manual Rule to force AU at OOBE for device pre M53 ( crbug.com/643408 ) ######################################################################## Rule { CohortId: 4 Update { UpdatedVersion: "8530.96.0" UrlBase: "http://dl.google.com/chromeos/parrot/8530.96.0/stable-channel" UrlBase: "https://dl.google.com/chromeos/parrot/8530.96.0/stable-channel" DefaultV2UrlBaseIndex: 1 PackageItem: { Name: "chromeos_8530.96.0_parrot_stable-channel_full_mp-v4.bin-013447fd5ba8cd3a397adc7bf4eac4db.signed" Size: "545728814" Hash: "JnI4B8DNd5qoI8vZEVr7YnVIOwg=" Sha256Hash: "f0a52bd24322cc66552d24dacfd8db49311241671d43333709f61a37c237b06d" } Pair { Tag: "ChromeVersion" Value: "53.0.2785.154" } Pair { Tag: "ChromeOSVersion" Value: "{updated_version}" } Pair { Tag: "MetadataSize" Value: "35152" } Pair { Tag: "MetadataSignatureRsa" Value: "Xcl7kOgwhv/mc7T7CpPoOhR8/9FxJHj/lpe166vRb3gZy20rZwjS7Q5BOfgKt9ULNVRbaLzVsgMjhiKV9OIJNWdnsltiC6iYJUFuqPVru6FvQlhzOvqJh96+4xOJh10+DmiBP48t5y4klceOvWCAXRRAsJpdmbjkIEfS+y96Y2I6ftx018kFs5dxYXfXpqwyp490VuPpu6niA34aSO3RLFlFo612+ukXWmAsVauOizHG0giK6Gz694AAPVnGM9zNJNUBA9dnizXX5wlriuCgfWVBC6abU11aXPIbEuxMWmxiTDi+3YDEetdFhKs5o/2WjXqjPCsdINYQb/CEEkO8/A==" } Pair { Tag: "sha256" Value: "8KUr0kMizGZVLSTaz9jbSTESQWcdQzM3CfYaN8I3sG0=" } Pair { Tag: "IsDelta" Value: "true" } Pair { Tag: "IsDeltaPayload" Value: "false" } Pair { Tag: "MaxDaysToScatter" Value: "14" } Pair { Tag: "deadline" Value: "now" } } Test { Version: "[2465.118.0-{updated_version})" } } Butterfly and stout also have this 53 stepping stone. I think we should be good to go already?