In the meanwhile I committed the fix as bug #592. You can just update from the repository.
Giovanni
[possible BUG] Bug 458 might not be fixed completely Topic is solved
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
Re: [possible BUG] Bug 458 might not be fixed completely
Because an error the driver is not functional if assertions are disabled, it will be fixed in 3.0.0p5.
Giovanni
Giovanni
-
- Posts: 15
- Joined: Mon Nov 25, 2013 11:10 am
Re: [possible BUG] Bug 458 might not be fixed completely
The patch Giovanni proposed certainly improved SDIO writes, I am running at SDC_CLK_50MHz again.
I'll let the system run for a while and feedback if I encounter any issues.
Thank you!
I'll let the system run for a while and feedback if I encounter any issues.
Thank you!
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
Re: [possible BUG] Bug 458 might not be fixed completely
Note that even if the card is recognized as 50MHz, the clock is always set to 25MHz, there is an STM32 errata preventing correct 50MHz operations.
Giovanni
Giovanni
Re: [possible BUG] Bug 458 might not be fixed completely
Hello,
I discovered this bug fix while doing the F7 USB-MSD benchmarks. I then found the "SDIO clock divider BYPASS" errata in F205/F207/F215/F217/F40x/F41x but it isn't mentioned anymore in the F7 errata (and maybe others), so I removed it. Maybe a condition could be added based on the MCU family to restore the 50MHz bypass ?
Thank you, Koen
I discovered this bug fix while doing the F7 USB-MSD benchmarks. I then found the "SDIO clock divider BYPASS" errata in F205/F207/F215/F217/F40x/F41x but it isn't mentioned anymore in the F7 errata (and maybe others), so I removed it. Maybe a condition could be added based on the MCU family to restore the 50MHz bypass ?
Thank you, Koen
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
Re: [possible BUG] Bug 458 might not be fixed completely
Hi,
Have you tested it on the F7?
Anyway, moving this topic in "bug reports".
Giovanni
Have you tested it on the F7?
Anyway, moving this topic in "bug reports".
Giovanni
Re: [possible BUG] Bug 458 might not be fixed completely
Yes and I haven't noticed a problem. I also tried your test code on page 3. F765VIT6.
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
- Giovanni
- Site Admin
- Posts: 14444
- Joined: Wed May 27, 2009 8:48 am
- Location: Salerno, Italy
- Has thanked: 1074 times
- Been thanked: 921 times
- Contact:
Re: [possible BUG] Bug 458 might not be fixed completely
I think this has been fixed already. If not, reopen it.
Giovanni
Giovanni
Re: [possible BUG] Bug 458 might not be fixed completely
Hello.
I'm using Chibios with STM32F437. This core is not affected by SD clock errata of older devices. Last days I made a lot of experiments with 48MHz SD clock - all seems fine for the moment. There is no such a issue in the errata sheet of the F437, too. The highest clock is important to achieve better performance.
So I think you can enable higher clock for F4 devices, or better include some macros to manipulate it on compile time.
I'm using Chibios with STM32F437. This core is not affected by SD clock errata of older devices. Last days I made a lot of experiments with 48MHz SD clock - all seems fine for the moment. There is no such a issue in the errata sheet of the F437, too. The highest clock is important to achieve better performance.
So I think you can enable higher clock for F4 devices, or better include some macros to manipulate it on compile time.
Who is online
Users browsing this forum: No registered users and 22 guests