Following is a report from inside of Google: **What did you do before the bug happened? (Steps to reproduce your issue)** ``` # With the attached patches, where all of the patches have a # Message-Id but the cover letter doesn't. git send-email *.patch ``` Specifically, you can see me doing it: ``` $ git send-email *.patch 0000-cover-letter.patch 0001-dt-bindings-interrupt-controller-arm-gic-v3-Add-quir.patch 0002-irqchip-gic-v3-Disable-pseudo-NMIs-on-Mediatek-devic.patch 0003-arm64-dts-mediatek-mt8183-Add-mediatek-gicr-save-qui.patch 0004-arm64-dts-mediatek-mt8186-Add-mediatek-gicr-save-qui.patch 0005-arm64-dts-mediatek-mt8192-Add-mediatek-gicr-save-qui.patch 0006-arm64-dts-mediatek-mt8195-Add-mediatek-gicr-save-qui.patch To whom should the emails be sent (if anyone)? Message-ID to be used as In-Reply-To for the first email (if any)? (mbox) Adding cc: Douglas Anderson <dianders@xxxxxxxxxxxx> from line 'From: Douglas Anderson <dianders@xxxxxxxxxxxx>' From: Douglas Anderson <dianders@xxxxxxxxxxxx> To: Cc: Douglas Anderson <dianders@xxxxxxxxxxxx> Subject: [PATCH 0/6] irqchip/gic-v3: Disable pseudo NMIs on Mediatek Chromebooks w/ bad FW Date: Thu, 11 May 2023 15:25:55 -0700 Message-ID: <20230511151719.6.Ia0b6ebbaa351e3cd67e201355b9ae67783c7d718@changeid> ``` If you look at `0000-cover-letter.patch` you can see that it has no Message-ID, but the above clearly shows that the cover letter is being sent with a Message-ID (and the one from the last patch). **What did you expect to happen? (Expected behavior)** The cover letter should get an auto-generated Message-Id like it always used to. **What happened instead? (Actual behavior)** The cover letter ends up with the same Message-Id as the last patch. **What's different between what you expected and what actually happened?** See above. **Anything else you want to add:** This happens when using the `patman` tool to send patches with a cover letter. The individual patches encode the "Change-Id" in the Message-ID but the cover letter doesn't need any special Message-Id. I didn't notice this until after I sent a patch series and thus: https://lore.kernel.org/r/20230511150539.6.Ia0b6ebbaa351e3cd67e201355b9ae67783c7d718@changeid/ ...refers to both my cover letter and the last patch in the series. **Please review the rest of the bug report below. You can delete any lines you don't wish to share.** ``` [System Info] git version: git version 2.40.1.606.ga4b1b128d6-goog cpu: x86_64 no commit associated with this build sizeof-long: 8 sizeof-size_t: 8 shell-path: /bin/sh uname: Linux 6.1.15-1rodete3-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.15-1rodete3 (2023-03-28) x86_64 compiler info: gnuc: 12.2 libc info: glibc: 2.36 $SHELL (typically, interactive shell): /bin/bash [Enabled Hooks] commit-msg pre-auto-gc pre-commit prepare-commit-msg ``` The report references an internal build based on 2.40.1, but we were able to reproduce on upstream git version 2.41.0.rc0.4.g004e0f790f (without using `patman`). Note that the internal report did come with the output of `git format-patch` as a zipped attachment, which I tried to attach and am hoping won't get filtered. As described in the report, the patches themselves already contain Message-IDs, but the cover letter does not. At the mail linked in the bug, you can notice that patch 6 has two Message-IDs for some reason, and the one that's clearly not generated by the `patman` helper is appended to patch 6 instead of patch 0. Bisecting shows that this bug was introduced at a8022c5f7b (send-email: expose header information to git-send-email's sendemail-validate hook, 2023-04-19), thanks Josh for bisecting it. - Emily
Attachment:
gitpatches.tgz
Description: application/compressed-tar
