TL;DR
Get the latest gadgets delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
A long-term GrapheneOS user reports that OsmAnd runs slowly on a Pixel 8 because of the operating system’s hardened memory allocator. Disabling that allocator for OsmAnd alone restored normal speed, at the cost of reduced exploit protection for that app.
A GrapheneOS user who ran the privacy-focused Android operating system on a Pixel 8 for about a year has traced a persistent performance problem in the OsmAnd mapping app to the operating system’s hardened memory allocator, and documented a per-app setting that restores the app’s speed. Writing on blog.wirelessmoves.com in September 2026, the author reported that OsmAnd ran noticeably slower on GrapheneOS than on other Android devices, and that switching off the hardened allocator for that single app made it “as fast again as ever.”
The author states that overall satisfaction with GrapheneOS’s security and privacy features was high after a year of use. The single recurring exception was OsmAnd (OpenStreetMap for Android), which the author observed to be noticeably slower on the Pixel 8 running GrapheneOS than on other Android devices. As a side effect of the slowness, the author began evaluating alternative map applications and found CoMaps, which according to the report “meets almost all of my needs,” though the author still wanted occasional access to OsmAnd.
After investigating, the author attributed the slowdown to GrapheneOS’s hardened memory allocator, a security feature that appears to create significant overhead for OsmAnd specifically. The author’s hypothesis is that this is linked to how map applications work: scrolling a map requires constant loading and discarding of data, a workload that repeatedly exercises the allocator. This explanation is the author’s inference rather than a finding from the GrapheneOS project itself.
The report’s key practical detail is that the hardening can be deactivated on a per-app basis. The documented steps: long-tap the OsmAND app icon, select ‘Info’, scroll down to ‘Exploit protection’, leave general protection enabled, and disable only the ‘hardened memory allocator’ toggle. According to the author, after restarting OsmAnd, the app was much faster. The author acknowledged a security drawback to this change but judged it acceptable for OsmAnd’s threat profile, noting the app does not load content from the internet apart from occasional map updates.
Performance Trade-Offs of GrapheneOS Hardening
The report highlights a tension at the core of GrapheneOS’s design philosophy. The hardened memory allocator exists to make memory corruption exploits harder to execute, and it is one of the features that distinguishes GrapheneOS from stock Android. But hardening typically carries a runtime cost, and this report suggests that cost can be severe for specific workload patterns — in this case, an app that continuously allocates and frees memory while a map is panned.
For users, the significance is that GrapheneOS exposes per-app control over exploit protection, allowing a granular compromise: keep full hardening everywhere except for the one app where the overhead is intolerable. This matters to anyone weighing a switch to GrapheneOS who is concerned that security hardening might degrade everyday responsiveness, and to OsmAnd users specifically, for whom the report offers an immediate, documented remedy.
The report also draws attention to CoMaps as a practical alternative within the OpenStreetMap ecosystem, which may be relevant to readers seeking map apps that perform well on hardened Android systems without any configuration changes.
GrapheneOS and the Hardened Allocator
GrapheneOS is a privacy- and security-focused mobile operating system based on the Android Open Source Project, most commonly installed on Google Pixel devices, which serve as the project’s reference hardware. The author’s test device, a Pixel 8, falls within that supported device family.
The hardened memory allocator referenced in the report is one of GrapheneOS’s exploit-protection measures, designed to make certain classes of memory-safety attacks more difficult. In the Android settings interface described in the report, it appears alongside general exploit protection under per-app app-info screens, meaning users can adjust it independently for each installed application.
OsmAnd is an open-source offline mapping application built on OpenStreetMap data, popular with users who want locally stored maps. Its rendering approach — loading and releasing map tiles and vector data as the user pans — is the workload the author suspects triggers the allocator’s overhead.
“There is one app, however, that runs noticeably slower on my Pixel 8 with GrapheneOS than on other Android devices: OpenstreetMap for Android (Osmand).”
— Author, blog.wirelessmoves.com
What the Single-User Report Doesn’t Prove
The findings come from a single user’s experience on one device, and no benchmark measurements are provided — the comparison between GrapheneOS and other Android devices is the author’s qualitative observation. The claim that the hardened allocator is the cause rests on the author’s before-and-after test after toggling the setting, not on profiling or an official statement.
The explanation for why the allocator affects OsmAnd — that map scrolling causes constant loading and discarding of data — is explicitly the author’s speculation. There is no comment in the source material from the GrapheneOS project or from OsmAnd’s developers on the issue, and no indication of whether other memory-intensive apps are similarly affected, whether the behavior varies by device or GrapheneOS version, or whether any upstream fix is planned.
Steps for Affected OsmAnd Users
Readers experiencing the same slowdown can follow the report’s documented procedure: open the app’s info screen via a long-tap on the OsmAnd icon, navigate to ‘Exploit protection’, and disable only the hardened memory allocator, then restart the app. Users should weigh the author’s stated caveat: turning the allocator off reduces that app’s exploit protection, a trade-off each user must judge against how much untrusted content the app handles.
Whether GrapheneOS developers address the underlying performance overhead, or whether OsmAnd adjusts its memory patterns, remains to be seen; the report contains no roadmap from either project. Readers who prefer not to change security settings could instead evaluate CoMaps, which the author reports meets nearly all mapping needs on GrapheneOS without configuration changes.
Key Questions
Why is OsmAnd slow on GrapheneOS?
According to a user report on blog.wirelessmoves.com, the slowness is caused by GrapheneOS’s hardened memory allocator, which adds overhead that appears to affect OsmAnd’s map-scrolling workload. This is the author’s diagnosis based on before-and-after testing, not an official finding.
How do I disable the hardened memory allocator for one app?
Long-tap the app icon, select ‘Info’, scroll to ‘Exploit protection’, and turn off only the hardened memory allocator option while leaving general protection enabled. Restart the app afterwards.
Is it safe to turn off the hardened memory allocator for OsmAnd?
It reduces that app’s protection against memory-corruption exploits. The report’s author judged the risk acceptable because OsmAnd loads almost nothing from the internet beyond occasional map updates, but each user should make their own assessment.
Are there alternative map apps that work well on GrapheneOS?
The same report identifies CoMaps as an OpenStreetMap-based alternative that met nearly all of the author’s needs, discovered while working around OsmAnd’s slowness.
Does this performance problem affect other apps or devices?
That is unclear. The report covers a single Pixel 8 and a single app, with no benchmarks and no statement from GrapheneOS or OsmAnd developers on broader impact.
Source: hn
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
