899352 - Increase the installation timeout for the stub installer

Status: RESOLVED FIXED

(Firefox :: Installer, defect)

Description

[Robert Strong (they/them - no direct email)]

Once case is that pave over installations will take longer and it looks like there pave over installs hit the installation timeout more often than non-pave over installations.

                            |   Timeout  |       All     |
----------------------------+------------+---------------+
Pave Over & Has Profile     |  57419, 3% | 2289001,  31% |
Pave Over & No Profile      |   4961, 6% |   81431,   1% |
Clean Install & Has Profile |  32288, 2% | 1295859,  18% |
Clean Install & No Profile  |  55454, 2% | 3624359,  50% |
All                         | 150122, 2% | 7290650, 100% |

We should be able to do this dynamically so the progress bar isn't wonky.

Comment 1

[Robert Strong (they/them - no direct email)]

Attachment: patch rev1

Comment 2

[Brian R. Bondy [:bbondy]]

Comment on attachment 8393876 [details] [diff] [review]
patch rev1

Review of attachment 8393876 [details] [diff] [review]:
-----------------------------------------------------------------

::: toolkit/mozapps/installer/windows/nsis/common.nsh
@@ +4358,5 @@
>      !verbose ${_MOZFUNC_VERBOSE}
>      !define OnStubInstallUninstall "!insertmacro OnStubInstallUninstallCall"
>  
>      Function OnStubInstallUninstall
> +      Exch $R2

Convention is to use decreasing order for register names, and to rename accordingly. I've done this in the past though and it's a pain so if you don't want to I won't hold up the review for it :)

Comment 3

[Robert Strong (they/them - no direct email)]

See the note for the macro
* When modifying this macro be aware that LineFind uses all registers except
* $R0-$R3 so be cautious. Callers of this macro are not affected.

I had to use $R2 and $R3 since $R0 and $R1 are used elsewhere and LineFind uses the rest.

Comment 4

[Robert Strong (they/them - no direct email)]

Pushed to mozilla-inbound
https://hg.mozilla.org/integration/mozilla-inbound/rev/b1500506a6c4

Comment 5

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/b1500506a6c4

Comment 6

[Robert Strong (they/them - no direct email)]

Comment on attachment 8393876 [details] [diff] [review]
patch rev1

[Approval Request Comment]
Bug caused by (feature/regressing bug #): None. This patch has the potential of improving stub installer successful install rates and I'd like to get metrics for this on the more heavily used channels sooner rather than later.
User impact if declined: None
Testing completed (on m-c, etc.): on m-c and manual testing
Risk to taking this patch (and alternatives if risky): small since it has been on m-c for several days and is fairly straightforward
String or IDL/UUID changes made by this patch: None

Comment 7

[Robert Strong (they/them - no direct email)]

Pushed to mozilla-aurora
https://hg.mozilla.org/releases/mozilla-aurora/rev/84dbc0d6ae90

Comment 8

[u279076]

Robert, is this something you need some QA for? If so, please advise how we can test the timeouts you refer to in comment 0.

Comment 9

[Robert Strong (they/them - no direct email)]

Anthiny, no QA needed for this.