#13600 closed bug (fixed)
WebPositive crashes on Intel Celeron
| Reported by: | Alexco | Owned by: | pulkomandy |
|---|---|---|---|
| Priority: | normal | Milestone: | R1/beta2 |
| Component: | Applications/WebPositive | Version: | R1/Development |
| Keywords: | Cc: | ||
| Blocked By: | Blocking: | ||
| Platform: | x86 |
Description
WebPositive crashes directly at startup on hrev51259.
Attachments (2)
Change History (15)
by , 7 years ago
| Attachment: | WebPositive-444-debug-11-07-2017-23-09-13.report added |
|---|
comment:1 by , 7 years ago
| patch: | 0 → 1 |
|---|
by , 7 years ago
comment:2 by , 7 years ago
| Platform: | All → x86 |
|---|
comment:3 by , 7 years ago
| Component: | - General → Applications/WebPositive |
|---|---|
| Owner: | changed from to |
comment:4 by , 7 years ago
| patch: | 1 → 0 |
|---|
comment:5 by , 7 years ago
comment:6 by , 7 years ago
Sorry, late reply. Understandable, but strange. System is usable, but only partially. Maybe include something that can at least view HTML files as replacement? Or a proper warning?
comment:7 by , 7 years ago
For viewing HTML on such old machines I would recommend installing NetSurf (no javascript support, however).
comment:8 by , 7 years ago
The "d"-suffixed XMM-register instructions were introduced with SSE2 and the Pentium 4 (but there is also an older zero-operand "movsd" instruction which is completely unrelated). If the OP's Celeron is at least from the Pentium III era, it has at least SSE.
BeZilla will work on it, although for security reasons I'd recommend disabling JavaScript and SSL 3.0 and enabling TLS 1.0.
I agree that there should at least be an error message if this isn't going to be fixed, as the official system requirements still don't mention what CPU features are required.
comment:10 by , 6 years ago
@PulkoMandy: There is a way to force JavaScriptCore to use the interpreter instead of the JIT, I believe, which should solve this issue. Maybe we could do that based on CPUID?
comment:11 by , 6 years ago
JIT is not the problem here. It's just that WebKit is developped mainly by Apple and they don't care about machines that old, so the default compiler args allow to use some modern CPU instructions. If it was only for Apple, they would have dropped 32bit support long ago already. For now Igalia has managed to convince them otherwise as they use WebKit in embedded contexts.
A look at the backtrace will show you that the crash is in the "currentTime" function, long before we are ready to start interpreting javascript code. I'm just not interested in setting up a PIII box just so I can test this. I may consider accepting patches that change the compiler flags...
comment:12 by , 6 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
We now block users from running WebPositive when SSE2 is not available, so this is "fixed".
comment:13 by , 5 years ago
| Milestone: | Unscheduled → R1/beta2 |
|---|
Assign tickets with status=closed and resolution=fixed within the R1/beta2 development window to the R1/beta2 Milestone
Your CPU does not support SSE. WebKit requires SSE. We already disabled SSE2 to allow it to run on Pentium 4 and later, but I don't think we want to do more efforts in that direction (as it would slow things down on more modern machines).