Ticket #8793: scanbuild-code-review.txt

File scanbuild-code-review.txt, 1.6 KB (added by andrea, 6 years ago)
Line 
1 * 08325b58bef83bfed181c493f269ef57477152c0 looks fine to me
2 * 41a8930fa1202b882687c0c3a328307b480934b5 strikes me as a good idea on
3   general defensive programming grounds anyway.
4 * d1be2f5cf8afc7d94f4c69081897d7ea3da71298 looks fine
5 * 0fd0f5f7a9309fb90a6a4d8bad7d6399a45c7cc1 looks like a definite win
6   - Could the bug it fixes ever arise under attacker control?
7 * 69ea4450caee65be56912fd2618c2b95413a0763 looks fine to me
8 * 710649257176a35b28c0ca5f2b823d39e011350c looks fine
9 * 7cd9520ba9713c10ef9f958a977a1d3d8d1a2c4c looks okay
10 * 895b6789e8b33af180a00d843eb25343bace4a4d looks good
11 * 78bc814c049fe67d0b521a288d3516248e110301 looks good
12 * 1b3bddd013dab6d0aa8159e1690d944e226ed77f looks okay
13 * d4ad254917ffa0dfba371624f72bc9e163645b8e looks okay
14 * 4d51dcda2fa75a3841e041ab7c3de325d73e2850
15   - But at least in theory we could have a hash table that large on a 64-bit
16     platform
17   - The numeric overflow would still be present because name##_PRIMES[]
18     would still be an unsigned int.
19 * 9c9e07963dddff6e11330e9dc8ad7a6d37da4aa4
20   - Someone tell scanbuild that if half the address space is full and
21     you malloc twice that size, the malloc is going to fail anyway.
22   - This patch looks okay but it overallocates slightly in the common
23     case.  Maybe malloc(len*2+ellipses+1) rather than malloc(len*2+4) ?
24 * 5670e38efb8529d3439b8a160e9f19c4147e01ad looks fine to me
25 * 1800e79ca508e555d43eb8ca36e9544f42c98944 looks fine to me
26 * 78f555a2480b03911e602c2c041a10fd010804b9 looks fine to me
27 * 685d450ab3823c578514ce6986d00c6e219abb43 looks good to me
28 * 3b1f7f75a7efa51ae5549a6413e90066cfe307a8 looks good to me