On Sun, Feb 14, 2021 at 1:56 PM Bin Meng <bmeng...@gmail.com> wrote: > > Hi Alexander, > > On Fri, Feb 12, 2021 at 5:25 AM Alexander Bulekov <alx...@bu.edu> wrote: > > > > On 210211 1154, Alexander Bulekov wrote: > > ... > > > I applied this along with <20210208193450.2689517-1-f4...@amsat.org> > > > "hw/sd/sdhci: Do not modify BlockSizeRegister if transaction in progress" > > > > > > I ran through the entire OSS-Fuzz corpus, and could not reproduce the > > > crash. > > > > > > Tested-by: Alexander Bulekov <alx...@bu.edu> > > > > > Hi Bin, > > Phil explained to me that this patch should fix the problem independent > > of > > "hw/sd/sdhci: Do not modify BlockSizeRegister if transaction in progress" > > Yes, that's what I expect too. > > > > > With only this patch, there are still crashes. Here are three > > reproducers: > > > > Some of these are quite long, so here are pastebins for convenience: > > Repro 1: https://paste.debian.net/plain/1185137 > > Repro 2: https://paste.debian.net/plain/1185141 > > Repro 3: https://paste.debian.net/plain/1185136 > > I will take a look.
I have figured out a fix and will send out for review and testing soon. Regards, Bin